Estimación inicial de recursos minerales de litio y rubidio en el proyecto Conchas.

Berkeley Energia Limited | LSE / ASX / BdM: BKY | ABN: 40 052 468 569 | www.berkeleyenergia.com SALAMANCA Ctra SA -322, Km 30, 37495, Retortillo, Salamanca, Españ a | T: +34 923 193903 PERTH Level 9, BGC Centre, 28 The Esplanade Perth, WA 6000 | T: +61 8 9322 6322 | F: +61 8 9322 6558

INFORMACIÓN PRIVILEGIADA

Berkeley Energia Limited (“Berkeley” o la “Sociedad”), en cumplimiento de lo previsto en el artículo 17 del Reglamento (UE) nº 596/2014 sobre abuso de mercado y en el 228 del Texto Refundido de la Ley del Mercado de Valores aprobado por el Real Decreto Leg islativo 4/2015, de 23 de octubre, mediante el presente escrito informa sobre la estimació n inicial de recur sos minerales de litio y rubidio en el proyecto Conchas.

Se adjunta a continuación el texto íntegro de nota informativa para conocimiento de los accionistas de la Sociedad.

En Salamanca , a 21 de julio de 202 6.

Ignacio Santamartina Aroca , representante, a efectos de notificaciones

Berkeley Energia Limited | LSE / ASX / BdM: BKY | ABN: 40 052 468 569 | www.berekeleyenergia.com SALAMANCA Carretera SA – 322, km 30, 37495 Retortillo, Salamanca , España | T: + 34 923 193 903 PERTH Level 9, 28 The Esplanade , Perth WA 6000 | T: +61 8 9322 6322 | F: +61 8 9322 6558

NEWS RELEASE | 21 July 2026

Maiden Lithium & Rubidium Mineral Resource Estimate for Conchas Project

Berkeley Energia Limited ( Berkeley or the Company ) is pleased to announce the maiden Mineral Resource Estimate ( MRE ) for its 100% owned Conchas Project ( Conchas or the Project ), as part of the Company’s Critical Minerals Exploration Initiative in Spain.

Highlights:

• Shallow, thick zones of lithium ( Li) and rubidium (Rb) mineralisation , with accessory tin ( Sn), caesium (Cs), ber yllium ( Be), niobium ( Nb) and tantalum ( Ta) hosted within a muscovitic leucogranite unit • Maiden MRE of 11.8Mt at 0.41% Li2O & 0.21% Rb 2O o Totals ~49,000t of contained Li2O & ~25,200t of contained Rb2O o 100% of MRE classified in the Inferred Mineral Resource category o Prepared by independent consultants, Maja Mining Limited, and reported in accordance with the JORC Code (2012 Edition) • Mineralisation from surface and remain s open at depth , supporting resource growth potential • MRE amenable to bulk -tonnage, open -pit mining • Preliminary metallurgical testing demonstrated very good recoveries of Li and Rb at acceptable grades using flotation and magnetic separation methods • MRE demonstrates Conchas hosts a globally significant Rb 2O resource • Rb is a critical raw material for advanced technology and industrial applications used in key sectors including defence and military, aerospace, communications, medical and renewable energy . The USA, Japan and NZ have both classified Rb as a Critical Mineral due to its strategic importance and growing demand in high -tech applications . It is also of strategic importance in Europe and Canada.

• Li is designated as both a Critical and Strategic Material for the European Union as batteries are central to decarbonisation, energy security, and industrial competitiveness .

Berkeley Executive Director, Mr Robert Behets, commented:

“This initial Mineral Resource Estimate is a significant milestone for the Company ’s Critical Minerals Exploration Initiative in Spain and clearly demonstrates the significant potential of the Conchas Project.

The MRE milestone follows two successful drilling campaigns and a positive initial assessment of the metallurgical characteristics of the multi -commodity mineralisation at Conchas which confirmed that very good recoveries of Li and Rb can be achieved using flotation and magnetic separation methods .

The resource commodity mix which includ es rubidium, a high value critical metal used in high -tech applications in key sectors such as defence and military , and lithium, a Critical and Strategic Material for the EU, highlights the potential strategi c impor tance of Conchas.

The Company will continue to systematic ally advance the Project’s development , with additional infill drilling to increase the resource classification , and a second phase of metallurgical test work to optimise the flotation and magnetic separation processes .” For further information please contact:

Robert Behets Francis co Bellón Executive Director Chief Operations Officer +61 8 9322 6322 +34 923 193 903

info@berkeleyenergia.com

2 Conchas Mineral Resource Estimate (MRE) Summary The m aiden Inferred MRE for Conchas totals 11.8Mt at 0.41% Li 2O & 0.21% Rb 2O for ~49,000t of contained Li2O and ~25,200t of contained Rb2O (above a US$100/t N et Smelter Return ( NSR ) cut-off).

The MRE is constrained by an open pit Whittle optimisation pit shell using appropriate mining and processing costs, processing recovery , and metal concentrate payability and revenue values .

The MRE has been p repared by independent consultants, Maja Mining Limited, and is reported in accordance with the JORC Code (2012 Edition) .

As per ASX Listing Rule 5.8 and the JORC reporting guidelines, a summary of the material information used to estimate the MRE is detailed below.

Project Location

The Conchas Investigation Permit ( Conchas IP ) covers an area of 31km2 in the southwest of the province of Salamanca , close to the Portuguese border (Figure 1). The Project is located ~21km from Ciudad Rodrigo and ~110km from Salamanca , in the municipalities of Espeja and Fuentes de Oñoro , and is readily accessible from established roads .

Figure 1: Conchas Location Plan Geology and Geological Interpretation

Regional Geology

The Conchas IP is in the central part of one of the most important and extensive geological zones of the Iberian Peninsula: The Central Iberian Zone. Li mineralisation is common in the Central Iberian Zone and, to a lesser extent, in the Galizia -Trás-Os-Montes Zone of Spain and Portugal, occurring along a ∼500km long north -northwest to south -southeast striking belt.

There are different styles of Li mineralisation along this belt, mainly associated with aplite -pegmatite bodies and, to a much lesser extent, with veins of quartz and phosphate. Li mineralisation is related to differentiation of peraluminous melts, which were generated by partial melting of metasedimentary rocks during the Variscan orogeny. Based on paragenesis and chemical composition, the pegmatitic fields and Li mineralisation associated with granitic cupolas record the highest fractionation levels.

3 There are a several textural and mineralogical indicators guiding Li exploration in the Central Iberian Zone and in the Galizia -Trás-Os-Montes Zone, including highly evolved peraluminous granites and pegmatites, the development of albite -rich and muscovite -rich assemblages, the occurrence of Li-bearing phosphates and silicates, and enrichment in incompatible elements that are commonly associated with LCT ( Li-Cs-Ta) pegmatite systems.

Project Geology

The Conchas deposit is situated within the Central Iberian Zone of the Iberian Massif which is characterised by pegmatite fields and Li enrichment in apical zones of highly fractionated granitic intrusions. The mineral occurrence is related to the outer rim of the eastern edge of the Guarda granitic batholith, where muscovite -

rich leucogranite hosts the elements of economic interest ( Rb, Li, Cs, Be, Ta, Sn, Nb) which have been explored for via the utilisation of geochemical , geophysical, and general geological exploration techniques of mapping, surface sampling and subsurface drilling (Figure 2).

Figure 2: Conchas Geolog ical Map The primary host rock is muscovite leucogranite, which is consistently associated with the observed mineralisation. Interbedded with this unit, particularly towards the western part of the deposit, are thicker and more frequent occurrences of regional granite, generally regarded a s barren material.

Six lithologies have been identified over the deposit from historical mapping and more recent drilling

activities :

• Mineralised Muscovitic Leucogranite ( LGM ) composed primarily of quartz, plagioclase, potassium feldspar and muscovite • Biotitic Porphyritic Granite - Coarse, Regional ( GMG ) composed primarily of quartz, biotite and feldspar • Biotitic Porphyritic Granite - Fine, Regional ( GMF ) composed primarily of quartz, biotite and feldspar • Muscovitic Granite ( GMO ) composed primarily of quartz, muscovite and biotite • Fine Grain Cupuliform Granite ( GCF ) composed of quartz, feldspar, muscovite and biotite • Quartz Dykes ( FQG ) Detailed core analyses reveal diverse mineralogical compositions. Muscovite, appearing as fine white flakes, is dispersed throughout the rock and often difficult to detect unless under reflecting light. Dark -coloured micas, likely biotite or possibly zinnw aldite, occur irregularly in clusters. Cassiterite has been positively identified within the LGM , and the presence of other minerals, such as rubellite or lepidolite, was inferred based on small, coloured crystals found in hydrothermally altered zones.

4 Structural features indicate some structural complexity. Contacts between the LGM and GMG are typically sharp and highly fractured, affecting both lithologies. Xenoliths of leucogranite are frequently observed in core, often displaying reddish alteration h alos.

Geological Interpretation

Since 2021, the Company has completed systematic exploration o n the Conchas IP, including soil sampling (2021, 2022), geological mapping (2024), geophysics (2025) , and drilling over two campaigns in 2022 (5 reverse circulation ( RC) holes for 282m) and 2024 (33 RC holes for 1,857m and 3 diamond ( DD) holes for 230m for metallurgical sampling), which has provided subsurface data with which to develop 2D and 3D geological and mineralisation interpretations which have fed into the evaluation of Mineral Resources.

Drilling activities have delineated sub -horizontal layers of the different lithotypes, with altered igneous basement ( CSG ), mineralised LGM and non -mineralised regional GMG the dominant lithologies modelled, along with zones of intercalated GMG/LGM (Figure 3).

Mineralised LGM vertical thicknesses range from 8m to 45m over southern and central portions of the deposit, thickening up to 70m to the north. The surface boundary of the LGM is defined based on the most recent (2025) geological mapping. This contact is generally steep, tending towards vertical, but with a slight dip inward.

Data collected over the Project to date supports the hypothesis of a multi -feeder system creating and controlling the spatial position of mineralised bodies as well as their thicknesses. An alternative interpretation posits the influence of tectonic processes, with blocks having b een uplifted or downthrown, leading to variations in erosion and preservation of mineralised material. Although direct evidence for such tectonic block movements is limited, the presence of highly tectonised zones and inferred fault structures lends some support to this model.

Figure 3: Conchas Cross Section

Drilling Techniques

RC an d DD drilling methods have been employed at Conchas to date.

RC drilling used a 140mm diameter face sampling hammer to collect a single sample for each metre drilled (Figure 4). Drillholes intercepted the mineralised horizons and up to 10m into barren regional granite. RC drilling rig utilised suitably sized compressors to ensure dry samples were collected, where possible. Plastic sample bags are strapped to the cyclone to maxim ise sample recovery. Sample logs record whether the sample is dry, moist or wet.

5 When the drillholes pass through zones with a significant water inflow, the sample bags are pierced to release the amount of water from the cyclone. In this operation, a part of the finest fraction is lost, resulting in a poorer recovery. However, no mater ial grade variation was observed across mineralised intervals that went from dry to wet.

DD drilling used to collect core sample for metallurgical testwork was of PQ (85.4mm core diameter) orientated using the REFLEXT ACT III tool. Due to challenging grounds conditions (hardness) core diameter was reduced to T101 (82.6mm core diameter) that did not allow for core orientation. HQ (63.5mm core diameter) was used at the end of hole. Drilling was via 3m drill runs.

The length of each run is measured directly on the core and compared with the start and end depths recorded by the drillers. Recovery for each run is calculated in a spreadsheet, along with the overall core recovery for each diamond drillhole. Core recover y exceeds 96%.

Figure 4: RC Drilling at Conchas (Drillhole CCR019)

Sampling Techniques

RC drill samples were collected at 1m intervals. RC intervals were sampled by splitting dry samples in the core shed to 3 –5kg using three -tier riffle splitter and further split to 0.7 –1kg using a one -tier riffle splitter.

Wet samples were split using a con e and quarter method.

Core samples of 5m intervals were sent for metallurgical testing. Core was cut in half using a diamond water saw, with care taken to ensure minimal ore loss. One half was placed into individually labelled bags numbered by metre, while the other half remain ed in the core boxes for future reference. Quarter core was submitted for metallurgical testwork.

6 Sample Analysis Methodology

Analysis

RC drill samples were sent to ALS Laboratory (Seville, Spain) for preparation and to ALS (Loughrea, Ireland) for analysis. Samples were dried, fine crushed down to 70% below 2mm and pulverised with at least 85% of the sample passing 75µm. 10g of sample was used for multi element analysis by the ALS ME-MS89L ™ Super Trace analysis method, considered appropriate for the style of mineralisation.

This method combines a sodium peroxide fusion with inductively coupled plasma -mass spectrometry ( ICP-

MS) instrumentation utilising collision/reaction cell technologies to provide the lowest detection limits available. A prepared sample (0.2g) is added to sodium peroxide flux, mixed well and then fused in at 670°C.

The resulting melt is cooled and then disso lved in 30% hydrochloric acid. This solution is then analysed by ICP-MS and the results are corrected for spectral inter -element interferences. The final solution is then analysed by ICP -MS, with results corrected for spectral inter -element interferences.

DD core samples were sent to Oviedo University (Asturias, Spain) and SLR laboratory (former Wardell Armstrong International , Cornwall, U nited Kingdom) for preliminary metallurgical testwork including crushing, grinding , head sample characterisation, scanning electron microscope ( SEM ) mineralogical analysis, gravimetry , flotation , and magnetic separation .

Quality Assurance and Q uality Control (QAQC ) A QAQC program was in place during all drilling activities to monitor accuracy and precision of analytical results, including the submission of field duplicate samples and blanks samples comprising 13% of total sampling (2022 campaign) and 9% (2024 campaign).

Monitoring for accuracy was only achieved via review of laboratory internal certified reference materials (CRM’s ), which show good correlation to certified values. External CRM’s were not submitted due to the inability to source appropriate element -specific CRMs for the mineralisation style and elemental suite. As such, there is some uncertainty in relation to the l aboratory accuracy. Additional measures are to be implemented in the next phase of drilling to ensure accuracy can be assessed.

For 2022 data, analysis of duplicate control plots suggests good correlation with correlation coefficients above 0.96 for Ta, Sn and Nb, and above 0.99 for Li, Rb and Cs.

Analytical results for 2022 blank samples show some instances of relative elevations and trends are noted, particularly for Li and Rb, and Cs and Nb. The Competent Person ( CP) notes that blank sample material is sourced from shales used in previous Company drilling campaigns, where low -level grades are detected.

The CP is of the opinion that the elevated grades observed are not significant and do not represent contamination. Ad ditional blank samples will be sourced for future drilling campaigns.

For 2024 data, quality control plots were generated using a ±10% deviation threshold between the duplicate and the original sample. Except for Sn, the duplicate results are consistent, with correlation coefficients above 0.98 and very few samples falling outside the defined limits. Relatively poor Sn correlation may be attributed to the irregular distribution of cassiterite and its higher nugget effect compared to the other elements.

The CP considers that analytical results show acceptable precision and that there has been no sample contamination.

Density

A total of 50 density determinations were taken from representative intervals within the three PQ DD holes.

Measurements were undertaken on samples representing the principal lithological units present within the resource area, including muscovitic leucogranite, altered muscovitic leucogranite, regional muscovitic granite and locally developed weathered material.

Dry bulk density measurements were determined using the Archimedes immersion method, whereby the dry mass of each sample was measured in air and subsequently weighed while submerged in water to determine sample volume by fluid displacement. This method is considered appropriate for competent, low -

porosity drill core samples.

7 To verify the accuracy and stability of the weighing system, three control measurements were completed for each drillhole; one at the start, one in the middle and one at the end of each hole. A total of nine control measurements were undertaken across the three drillholes. The control measurements demonstrated consistent performance of the weighing system throughout the density determination program.

The measured density dataset returned largely consistent results for the dominant lithologies hosting mineralisation. Average dry bulk density values of 2.644g/cm³ and 2.643g/cm³ were obtained for the muscovitic leucogranite and regional granite respective ly. Altered muscovitic leucogranite returned a slightly lower average density of 2.529g/cm³, while completely weathered material returned an average density of 1.575g/cm³.

The density data demonstrate that the principal mineralised lithologies display remarkably consistent density characteristics, with average values effectively identical between the muscovitic leucogranite and regional granite units. These lithologies accou nt for the vast majority of the mineralised volume incorporated within the MRE. Completely weathered material represents only a minor component of the geological model and is largely restricted to shallow near -surface zones.

Data Verification

The CP, visited the Project between 15 January and 17 January 2026 during which time the following activities were completed:

• Discussion s with technical personnel and received presentations related to geological setting and exploration activities completed by the Company • Ground truthing and visits to drillhole collar locations • Review of hand specimens collected over the Project and review of RC chips (holes CCR -005, CCR -

025)

• Inspection of drill core material and review of geological logging information (holes CCD -007, CCD -

021, CCD -028)

• Inspection of storage facilities containing RC sample rejects and pulp samples.

The CP reviewed sample preparation, analysis and security protocols and procedures used by Berkeley and sighted RC samples and chip trays during the site visit. The sampling and assaying QAQC program in place during the drilling campaigns was reviewed, including procedures, QAQC datasets, plots and graphs and spot checks undertaken of digital assay data against laboratory certificates.

It is the opinion of the CP that the adequacy of the sample preparation, security and analytical procedures and the results of the QAQC program suggest data is of sufficient confidence, particularly precision, for downstream use in Mineral Resource evaluation. However, accuracy cannot be assessed since external standard reference material was not included in the sample stream since releva nt standards could not be sourced. A review of internal laboratory QAQC suggests no issues in relation to accuracy and precision confidence.

Estimation Methodology

The drillhole database, relevant surface digital terrain models , and geology volume models were supplied by the Company . The elements estimated in the MRE were Be, Cs, Li, Rb (recovered in a Mica float) and Nb, Ta and Sn (recovered as gravity concentrate). The CP determined appropriate mineralisation domains, completed statistical analysis, grade estimation, reasonable prospects for eventual economic extraction (RPEEE ) and mineral resource classification based on the guidelines defined in the JORC Code (2012).

Data used for M RE Sub-surface drilling data and surface geochemical data was used in the interpretation of geology domains.

Digital wireframe surfaces and wireframe volumes generated by the Company were reviewed and modified as appropriate for the purposes of Mineral Resource evaluation. Finalised domains include the topography surface digital terrain model ( DTM ), base of CSG, mineralised LGM and the regional GMG. Ni ne internal GMG waste zones were modelled within the LGM.

8 Due to the number of potential saleable elements hosted in the muscovite leucogranite, a conceptual net smelter return ( NSR ) value was estimated from the in -situ grade of the seven elements of interest. Revenue, process recovery and element saleable percentages , and operating costs were estimated by the Company and are considered conceptual.

For Li, Rb, Cs and Be, the NSR model assumes recovery into a mica concentrate as the saleable product.

Given the early -stage nature of the Project and the absence of established commercial terms for such a concentrate, conservative saleability assumptions were applied to the contained metal value. These assumptions were adopted to provide a prudent basis for Mineral Resource reporting and conc eptual pit optimisation and are considered appropriate for demonstrating RPEEE.

All elements were converted to their saleable oxide molecules. A total NSR including partial NSRs w as estimated. The partial NSRs were estimated based on the mica concentrate ( N_Mica ) and heavy mineral gravity concentrate ( N_HM ).

A selective compositing process ( CompSE ) was run for both N_Mica and N_HM NSR values at a range of NSR cut -off values. This process was used to determine the spatial continuity both downhole and between drillholes to define coherent zones of mineralisation based on a range of NSR cut -off values .

Internal waste zones (gaps in the CompSE composites) were interpreted and estimated as a separate domain.

Using a combination of the geology, mineralisation, internal waste wireframes and the topography DTM , a 3D block model was constructed (Figure 5). A boundary defining the limits of mineralisation based on 80m influence from the nearest drillhole (80m is roughly equivalent to the average drill spacing). A constraining waste boundary based on the maximum possible open pit dimensions was also interpre ted to reduce the overall size of the final block model.

Figure 5: 3D Perspective View looking NW showing Volume Block Model

9 Drillhole samples were coded with ROCK and the Mica and HM mineralisation domains. The process was designed so that the Mica and HM domains could be independent if considered necessary. Analysis of both the spatial distributions and correlation between the elements led to the decision to use the Mica domain boundary for both Mica and HM mineralisation domains.

All samples are 1m length. Any non -assayed intervals downhole have been set to absent grades, which means the results from the neighbouring drillholes will be used to inform the grades proximal to the missing data.

Population statistics were reviewed for all grade estimation domains, including top capping requirements which were required as spatially isolated high -grade outliers are present. This prevents local overestimation of high grades and is particularly import ant with wide spaced mineral resource drilling, as the local overestimation can have a material impact on mine planning and project economics.

Mineralisation is hosted in two product groups – Mica, containing Rb, Cs, Li and Be; and Heavy Minerals (HM) made up of Sn, Ta and Nb. Both groups generally fall within the same common spatial mineralisation domain.

The Mica group currently has the greatest economic value, so the Mica domain has been used to define the main zones (North and South) of economic interest for both groups.

Rb has the best -behaved lognormal distribution within the Mica group with a very low coefficient of variation.

Both Li and Cs show a good correlation with Rb, with Be having a very low correlation with Rb.

Rb was used for the variography modelling for the Mica group, with Li and Cs being linked to Rb for grade estimation to maintain the grade correlation between the three major Mica elements. Be was treated independently.

Nb was used for the variography modelling for the HM group, with Sn and Ta being linked to Nb for grade estimation.

Variograms were modelled for Rb, Be and Nb to determine spatial continuity parameters for grade estimation using ordinary kriging ( OK).

Nugget effect for both Mica domains is around 15 –20% which indicates that the area closest to each sample will be strongly influenced by the local grade, i.e. less smoothing required and reasonable reliability for local grade estimates. The nugget for the HM domains at around 5% demonstrates very good confidence in local grade estimates.

In the Mica group, around 60% of the deposit variability occurs with a 60 –100m range, demonstrating that a drill pattern of around 40m spacing will produce a high confidence grade estimate. Note that the grade continuity of the HM group elements is slightl y higher than the Mica group elements.

Generally, the Northern domain has slightly better continuity than the Southern domain.

The shape and nature of the variograms combined with the reasonably well-behaved sample grade populations support the use of OK for grade estimation.

All block grade estimates were completed using OK for the main mineralisation domains and internal waste/low -grade domain. The surface CSG, bulk waste LGM and bulk waste regional CMG units were all estimated using inverse distance to the power of 2 ( ID2) due to lack of adequate sample data to model variograms and the fact that these zones are not part of the MRE but will be required for mine planning.

Grades were estimated into parent blocks with a panel size of 40m (N) by 40m (S) by 5m (Elev). Block discretisation using 2 by 2 by 5 was applied to the estimation panels. No octant search parameters were used as OK is the best declustering algorithm and successfully resolves any sample clustering issues within the given search ellipse.

The block model was validated both visually and statistically .

10 Swath plots were calculated which show the comparison between model grades and sample grades within a grid, horizontal, east -west and south -north. Global comparison of average domain sample grades with block model grades is excellent, with no apparent bias shown in the block model grade estimate.

Visual review shows expected grade trends when compared to the sample data.

The measured density dataset returned largely consistent results for the dominant lithologies hosting mineralisation. Average dry bulk density values of 2.644g/cm³ and 2.643g/cm³ were obtained for the muscovitic leucogranite and regional granite respective ly. Altered muscovitic leucogranite returned a slightly lower average density of 2.529g/cm³, while completely weathered material returned an average density of 1.575g/cm³.

The density data demonstrate that the principal mineralised lithologies display remarkably consistent density characteristics, with average values effectively identical between the muscovitic leucogranite and regional granite units. These lithologies accou nt for the vast majority of the mineralised volume incorporated within the MRE.

Completely weathered material represents only a minor component of the geological model and is largely restricted to shallow near -surface zones. Given its limited volumetric contribution to the mineralised domains, the lower density values associated with weathered material are not considered material to the global MRE.

Mining and Metallurgy Factors

Metallurgical Testwork

Two studies investigating mineralogy and metallurgical response have been previously completed and disclosed on Conchas sample material :

• Mineralogy (2023) • Scoping Level Metallurgical Testwork (2025)

Mineralogy

In 2023 , six representative samples were sent to the Universidad del País Vasco, Department of Geology for mineralogical assessment. Samples were selected from outcrops over Conchas. For each sample, a 30μm thin section was prepared for subsequent petrographic study using a Leica DMPL polarising light microscope with an attached digital camera. The chemical analyses of thin section minerals were carried out at the Scientific -Technical Services of the University of Oviedo, using a CAMECA SX -100 electron microprobe (equipped with five WD S spectrometers, an energy dispersive spectrometer and SE, BSE and cathodoluminescence detectors). Petrography was completed on four samples. 182 mica analyses (129 analyses of leucogranite samples and 53 analyses of greisen samples) and 68 feldspar analyse s (on leucogranite) was completed. The study concluded :

• Among the mineral phases identified, micas are the only ones containing Li • The higher the iron (Fe) content, the higher the fluorine (F) content and, therefore, the higher the Li content. In turn, the micas with high F content are also those with the highest Cs and Rb content • The secondary phosphates identified are often derived from the alteration of primary phosphates containing Li-Fe-Mn in their composition, so that in unaltered leucogranite samples some phosphates may also contain Li in their composition Scoping -Level Metallurgical Testwork SLR Consulting Ltd ( SLR) was engaged to undertake scoping -level metallurgical testing on a composited sample of Rb-Li mineralisation from the Conchas deposit in 2025 .

The composite consists of 15 samples from the three DD holes drilled in the 2024 campaign. From each drillhole, five samples were selected, each composed of quarter core, with sample weights ranging between 2.57kg and 3.99kg. The total weight was 47.63kg.

The program of work included head sample characterisation , SEM mineralogy , gravity testwork, flotation testwork and magnetic separation testwork.

11 The sample was crushed to 100% passing 2mm blended and representatively split into 2kg sample lots for the test program. A representative subsample was split out of one of the 2kg samples, pulverised to 80% passing 75μm and split into four equal subsamples. One of the subsamples was submitted to SLR in -house analytical laboratory for head assay and analysed for Li, Rb, Fe, Ta, Nb, Be, and Sn. In addition to the in -

house head assay, a representative subsample was also submitted to ALS Global for detailed multi -element analysis.

A representative subsample of the -2mm feed material was subjected to particle size analysis by screen.

A 2kg sample was milled to a nominal D 80 of 300µm and sized to generate five fractions for size -by-size analysis and subsamples for mineralogical investigation.

Representative subsamples of the fractions were pulverised and submitted for Li, Rb, Ta, Nb, Be, Sn, Fe, cerium (Ce) and Cs assay.

The target mineral phases identified by SEM mineralogy analysis include cassiterite, Nb-Ta oxides, polylithionite and muscovite. Muscovite was the most abundant target phase, maintaining relatively consistent concentrations across all size fractions. Cassiterite and Nb-Ta oxides were both present in trace quantities. Polylithionite showed overall poor liberation and muscovite exhibits moderate liberation overall.

Gravity release analysis (GRA ) was completed on four fractions in the +11 µm size ranges generated. The results show that cumulative Li recoveries into the combined concentrates and middling product ranged from 28.0% at a grade of 0.16% Li ( -53 +11 µm) to 65.8% at a grade of 0.24% Li in the -150 +53 µm fraction.

Cumulative Rb recoveries into the combined concentrates and middling product ranged from 22.2% at a grade of 2,358ppm Rb (+300µm) to 66.1% at a grade of 2,049ppm Rb (2,242ppm Rb 2O) in the -150 +53µm fraction.

The results showed optimum liberation size for the Conchas composite was in the -150 +53µm fraction.

A short programme of flotation testing was performed on the Conchas composite to evaluate potential grades and recoveries at two grind sizes.

Two rougher tests were conducted at the 300µm (FT1 -300) and 150µm (FT2 -150) primary grind sizes to identify the better flotation performance, and one cleaner test was then conducted at the better performing grind size to evaluate the effect of kinetic clea ning on grades and recoveries.

The results of the rougher tests confirmed that the finer 150µm grind was the better performing test and was therefore used for cleaner flotation testing (FCT1 -150). Cleaner flotation achieved 87.2% Li stage recovery, representing 77.5% overall recovery (a fter desliming) at a grade of 1.04% Li (2.23% Li 2O), 70.9% Rb stage recovery representing 62.7% overall recovery at a grade of 0.79% Rb (0.87% Rb 2O), and 78.5% Cs recovery at a grade of 661ppm Cs.

Flotation testing of the Conchas material demonstrated that very good recoveries of target minerals could be achieved at acceptable grades Representative sub -samples of the 300µm and 150µm primary grinds were subjected to magnetic separation testing to evaluate potential grades and recoveries at the two grind sizes.

The 300µm sub -sample was screened at 150µm and the two fractions treated separately. The +150µm fraction was treated on an Eriez Log 1.4 -disc separator, the -150µm treated on a Bunting Wet High Intensity Magnetic Separator ( WHIMS ) 500 jaw magnetic separator and the results combined to generate the overall performances. The 150µm sub -sample was treated on the Bunting WHIMS 500 jaw magnetic separator.

The initial magnetic test intensity was 4,000 Gauss with testing conducted in 1000 Gauss increments up to 15,000 Gauss.

Magnetic separation testing on the -300µm +150µm material, which accounted for 36.4% of the mass, showed 76.6% of the Li (32.2% overall) and 57.7% of the Rb (23.4% overall) reporting to the magnetic product grading 2.34% Li 2O and 0.73% Rb. This result may present an opportunity for magnetic separation processing of a coarser +150µm fraction followed by flotation of the finer -150µm material.

12 Magnetic separation on the -300µm +150µm material also showed 43.5% of the Ta and 50.9% of the Nb reported to the combined 4,000, 6,000 and 9,000 Gauss magnetic concentrates grading 1,161ppm Ta and 1,551ppm Nb.

Metallurgical testwork completed in 2025 (a scoping -level test program) demonstrates good recoveries for the economic elements of interest, at acceptable grades using flotation and magnetic separation methods.

RPEEE

RPEEE is based on the Project demonstrating a positive conceptual financial return based on the production of two groups of saleable products. The primary product is a Mica float concentrate containing saleable Li, Rb, Cs and Be. A secondary and less valuable product stream is a HM gravity concentrate containing saleable Ta, Sn and Nb.

The deposit is on or close to the surface and is currently drill tested to around 150m depth in the eastern side of the northern mineralisation domain, so open pit mining is planned. Three DD drillholes we re completed for metallurgical testwork. A representative sample was composited and submitted to SLR for metallurgical testwork . The work included: 1) sample characterisation; 2) SEM mineralogy; 3) gravity testwork; 4) flotation testwork; and 5) magnetic separation testwork.

Based on the metallurgical testwork and peer review of product markets and potential revenue, an updated NSR algorithm was compiled by mining consultants, Mining Sense Global SL (Mining Sense ) based on a 5%, 10% and 50% payability for the elements in the Mica concentrate. Mining and processing costs are accounted for in the Whittle open pit optimisation process so are not included in the NSR calculation. A conceptual breakeven NSR of $18/t was calculated.

The conceptual pit optimisation utilised the most conservative of the NSR scenarios evaluated. A 5% payability assumption was selected to reflect the current uncertainty associated with future concentrate marketing arrangements and to provide a conservative basis for demonstrating RPEEE. Sensitivity analysis indicates that high er payability assumptions result in only limited changes to the optimised pit shell.

Rb is a critical raw material with growing significance in advanced technology and industrial applications, including in the defence and military, aerospace, communications, biomedical and renewable energy sectors. C s is used in the same market space and can be replaced by Rb in many applications, where Cs supply is unavailable or restricted. Both metals are indispensable for producing special crystals used in night -vision equipment and fibre -optic telecommunications systems. Other applications include precision timekeeping in atomic clocks, which are vital for GPS, telecommunications, and space exploration. Li is designated as both a Critical and Strategic Material for the European Union as batteries are central to decarbonisation, energy security, and industrial competitiveness .

Although the market for both Rb and Cs is limited, demand is expected to grow at 3 –5% per annum. The long-term outlook for Li demand remains strong . At current market revenues, the calculated breakeven NSR is US$18/t. However; Conchas does have two significant high -grade zones with good mining continuity at the US$100/t NSR cut -off. To reduce the risk of potential market over supply and any future decr ease in product revenue, reporting the MRE at the higher NSR cut -off of US$100/t is considered prudent.

Whittle open pit optimisation was used to determine the final RPEEE constraining open pit shell, with sensitivity runs completed based on different concentrate metal payability assumptions. Until a bulk concentrate can be produced and tested on various markets, the payability of the mica -hosted metals is the biggest unkno wn for the project economics. Whittle runs were completed using 5%, 10% and 50% payability.

There is no difference from 10% to 50%, with both open pit shell extracting all available mineralisation. At 5% payability, there is a 40% reduction in mineralisati on at the nominal NSR cut -off of US$18/t.

Due to the uncertainty in the mica concentrate payability a conservative 5% factor has been selected to satisfy RPEEE for this maiden MRE. However; once additional process and marketing studies have been completed any increase in the payability value may r esult in a material increase in the MRE.

Parameters used for the Whittle open pit optimisation runs are summarised Appendix B.

13 Cut-Off Grade At current market revenues, the calculated breakeven NSR is US$18/t. However; Conchas does have two significant high -grade zones with good mining continuity at the US$100/t NSR cut -off (Figure 6). To reduce the risk of potential market over supply and any future decrease in product revenue, reporting the MRE at the higher NSR cut -off of US$100/t is considered prudent.

Figure 6: 3D Perspective View looking NW of MRE blocks >=US$100/t NSR, within RPEEE optimised pit shell (with d rillhole collars ) Mineral Resource Classification The MRE is has been classified based on the guidelines presented in JORC Code (2012 Edition). The whole MRE has been classified as Inferred for the following reasons:

• Adequate geological understanding based on surface mapping, sampling and drilling • Adequate grade continuity and definition of grade domains based on supporting drilling, sampling and assay data with a NSR equivalent for all seven potentially recoverable metals • QAQC controls on all drilling, sampling and assaying procedures, appreciating that laboratory accuracy cannot be demonstrated since no external CRMs are included in the QAQC program • Adequate in -situ dry bulk density estimation to convert volume to dry metric tonnes • Adequate mining and metallurgical testwork to demonstrate the potential mining and processing methods for recovery of both a mica concentrate and gravity concentrate with saleable metal

concentrations

• Good standing of the existing tenement package with appropriate technical support to manage the ongoing advancement of the Project to satisfy legal, social and environmental requirements to develop a mining project in Spain • None of the MRE has been classified as Indicated or Measured as additional drilling, bulk density and metallurgical data is required to improve geological, grade, tonnage, recovery and marketing confidence.

The MRE is reported using a NSR breakeven cut -off of US$100/t with all material being constrained to the RPEEE open pit shell.

14 MRE Table The maiden MRE for Conchas is reported as of Ju ly 2026 and at a NSR cut -off of US$100/t (Table 1) .

RPEEE have been satisfied through the application of appropriate revenue, mining and processing parameters.

Table 1: Conchas MRE (July 2026 ) Conchas Project - MRE (July 2026 ) Classification Domain Tonnes

(Mt) NSR5 Rb 2O

(ppm) Li2O

(ppm) Cs2O

(ppm) BeO

(ppm) Ta2O5

(ppm) SnO 2 (ppm) Nb 2O5

(ppm) Dry

BD Inferred North LGM HG 4.14 112 2,061 3,898 132 168 62 760 95 2.64

South LGM

HG 7.69 116 2,165 4,265 134 134 64 727 96 2.64

Total 11.83 114 2,128 4,137 133 146 64 739 96 2.64

Notes:

• MRE includes all resource model blocks >= US$100/t NSR based on 5% payability of metals within the mica float concentrate.

• MRE is constrained by an open pit Whittle optimisation pit shell using appropriate mining and processing costs, processing recovery and metal concentrate payability and revenue values.

• Mineral Resources are not Mineral Reserves – further detailed economic studies and additional modifying factors are required.

• The MRE was classified following the guidelines presented in JORC Code 2012. The CP is Malcolm Titley of Maja Mining Limited.

• Totals may not add up due to rounding.

The NSR cut-off and tonnage relationship within the RPEEE constraints based on a breakeven NSR cut -off of US$18/t is present ed in Table 2.

Table 2: Conchas NSR Cut-off Grade -Tonnage Tabulation

NSR5

cut-off Tonnes

(Mt) NSR5 Rb 2O

(ppm) Li2O

(ppm) Cs2O

(ppm) BeO

(ppm) Ta2O5

(ppm) SnO 2 (ppm) Nb 2O5 (ppm) Dry BD 100 11.83 114 2,128 4,137 133 146 64 739 96 2.64 90 15.39 110 2,048 3,832 128 139 62 717 94 2.64 80 17.97 106 1,985 3,623 125 136 60 694 92 2.64 70 23.54 99 1,859 3,143 116 132 56 629 87 2.64 60 33.45 89 1,708 2,615 104 120 50 531 80 2.64 50 35.81 87 1,672 2,515 101 115 48 513 78 2.64 40 36.33 86 1,661 2,489 100 115 48 509 78 2.64 30 37.16 85 1,639 2,449 99 114 47 501 77 2.64 18 57.00 63 1,234 1,733 76 81 32 349 57 2.64

Next Steps

The Company will continue to systematically advance the Project’s development, with additional infill drilling to increase the resource classification, and a second phase of metallurgical test work to optimise the flotation and magnetic separation processe s.

15 Competent Persons Statements The information in this announcement that relates to Mineral Resources is based on, and fairly represents, information compiled by Mr Malcolm Titley, a Competent Person who is a member of The Australasian Institute of Mining and Metallurgy and member of Australian Institute of Geoscientists. Mr Titley is the principal of Maja Mining Limited , an independent consulting company . Mr Titley has sufficient experience that is relevant to the style of mineralisation and to the activity being undertaken to qualify as a Competent Person as defined in the 2012 Edition of the ‘Australasian Code for Reporting of Exploration Results, Mineral Resources and Ore Reserves’. Mr Titley consents to the inclusion in the announcement of the matters based on their information in the form and context in which it appears.

The information in this announcement that relates to Exploration Results including Metallurgical Test Work is extracted from announcement s dated 29 January 2025, 28 October 2025 and 29 April 2026 , which are available to view at www.berkeleyenergia.com . Berkeley confirms that: a) it is not aware of any new information or data that materially affects the information included in the original announcement; b) all material assumptions and technical parameters underpinning the Exploration Results in the orig inal announcement continue to apply and have not materially changed; and c) the form and context in which the relevant Competent Persons’ findings are presented in this announcement have not been materially modified from the original announcement.

Forward Looking Statements Statements regarding plans with respect to Berkeley’s mineral properties are forward -looking statements.

There can be no assurance that Berkeley’s plans for development of its mineral properties will proceed as currently expected. There can also be no assu rance that Berkeley will be able to confirm the presence of additional mineral deposits, that any mineralisation will prove to be economic or that a mine will successfully be developed on any of Berkeley mineral properties. These forward -looking statements are based on Berkeley’s expectations and beliefs concerning future events. Forward looking statements are necessarily subject to risks, uncertainties and other factors, many of which are outside the control of Berkeley, which could cause actual results to differ materially from such statements. Berkeley makes no undertaking to subsequently update or revise the forward -looking statements made in this announcement, to reflect the circumstances or events after the date of that report.

This announcement has been authorised for release by Mr Robert Behets, Director.

16 Appendix A: About Rubidium1,2,3,4,5 Rb is a critical raw material with growing significance in advanced technology and industrial applications, including in the defence and military, aerospace, communications, biomedical and renewable energy sectors.

Its unique properties make it indispensable for producing special crystals used in night -vision equipment and fibre -optic telecommunications systems. Other applications include precision timekeeping in atomic clocks, which are vital for global positioning systems ( GPS ), telecommunications, and space exploration.

Rb compounds play a key role in the production of specialty glasses, cutting -edge electronics, radiation detection devices and medical imaging technologies, ensuring their relevance across multiple high -growth sectors.

Specialty glasses, currently the largest market for Rb, are utilised in night vision equipment and fibre -optic telecommunications systems. Rb carbonate is used as an additive to these types of glass, lowering electrical conductivity and improving stability and durability.

Rb's photo -emissive properties lead to its application in motion -sensor devices, night -vision devices, photoelectric cells, and photomultiplier tubes. These applications highlight its importance in advanced electronic devices, particularly in sectors requiri ng precision and reliability.

Its application in photocells, which convert light into electric currents, is significant. These photocells are primarily used as sensors to regulate lighting in buildings, showcasing Rb's role in energy -efficient technologies.

Rb-based atomic clocks are used in military communication systems, navigation equipment, and precision -

guided weapons. The increasing focus on defence modernisation and the need for secure and reliable communication systems are expected to drive the demand f or Rb in the military sector.

Rb is also increasingly used as a key component in advanced batteries, particularly in the development of high-energy -density batteries for electric vehicles and renewable energy applications.

Global production of Rb is limited, with no Rb production recorded globally outside of China in 2023.

Due to its strategic importance and growing demand in high -tech applications used in key industry sectors, the United States of America and Japan have both classified Rb as a Critical Mineral, essential to their economic or national security, and with a supply chain vulnerable to disruption.

References

1 www.mordorintelligence.com/es/industry -reports/rubidium -market 2 www.straitsresearch.com/report/rubidium -market 3 www.marketresearchfuture.com/reports/rubidium -market -27298 4 U.S Geological Survey, Mineral Commodity Summaries, January 2024 - Rubidium 5 www.usgs.gov/news/national -news -release/us -geological -survey -releases -2022 -list-critical -minerals

17 Appendix B: JORC Code, 2012 Edition – Table 1 Section 1 Sampling Techniques and Data (Criteria in this section apply to all succeeding sections.) Criteria JORC Code explanation Commentary

Sampling

techniques Nature and quality of sampling (eg cut channels, random chips, or specific specialised industry standard measurement tools appropriate to the minerals under investigation, such as down hole gamma sondes, or handheld XRF instruments, etc). These examples should not be taken as limiting the broad meaning of sampling. Berkeley Energia Limited (Berkeley or the Company) completed drilling campaigns in 2022 (5 reverse circulation (RC) holes for 282m) and 2024 (33 RC holes for 1,857m and 3 diamond (DD) holes for 230m for metallurgical sampling)

RC drilling

RC drill samples are collected over 1 m intervals. All samples are placed into individually labelled, consecutively numbered sample bags. The RC samples obtained are considered representative of the material drilled.

DD drill core was not used for routine resource assaying. Selected intervals were sampled for metallurgical testwork and density determination.

The selected core intervals have been cut in half, with one half placed into individually labelled bags numbered by the metre , while the other half will remain in the core boxes for future reference. The DD samples obtained are considered representative of the material drilled.

Metallurgical Drilling

The metallurgical testwork sample is a composite of samples obtained from three DD drillholes completed in 2024. The locations of the three DD drillholes within the Conchas Investigation Permit (IP)

follow:

• South – drillhole ID: CCD007 (refer ASX Announcement dated

29/01/2025)

• North -East – drillhole ID: CCD021 (refer ASX Announcement

dated 29/01/2025)

• North -West – drillhole ID: CCD028 (refer ASX Announcement

dated 29/01/2025)

The sampling strategy is guided by the criterion of proximity to RC drillholes for which analytical data are available, thereby facilitating the most reliable estimation of expected sample grades.

Five consecutive samples from each DD drillhole were selected, with each sample corresponding to a 1 m interval. The intervals chosen

were:

• CCD007: from 45 m to 50 m • CCD021: from 65 m to 70 m • CCD028: from 15 m to 20 m Each metre of core was cut into two halves, and one of these halves subsequently divided into quarters. From the two quarters produced, a single quarter (consistently the same one) was selected to obtain the individual samples. These quarter -core samples w ere placed into bags clearly labelled with the DD drillhole ID and depth interval and sealed with cable ties.

The weight of individual samples ranged from 2.6 kg to 4.0 kg. The total weight per DD drillhole interval was:

• CCD007: 15.6 kg • CCD021: 16.3 kg • CCD028: 15.5 kg The overall mass of material submitted to the laboratory for metallurgical testing was 47.4 kg.

Include reference to measures taken to ensure sample representivity and the appropriate calibration of any measurement tools or systems used. RC Drilling Field duplicate and blanks samples are inserted into the sample stream to assess the variability of mineralisation. Approximately 13% of all samples relate to quality control. In addition, the laboratories

18 Criteria JORC Code explanation Commentary undertake their own duplicate sampling as part of their internal quality assurance/quality control (QAQC) processes.

Examination of the QAQC sample data to date indicates satisfactory performance of field sampling protocols and assay laboratories providing acceptable levels of precision and accuracy.

The survey of drillhole collar locations has been completed by a qualified surveyor using standard differential global positioning system (GPS) equipment achieving sub decimetre accuracy in horizontal and vertical position.

Downhole surveys have been completed for 34 of the 36 holes drilled in the 2024 campaign. Downhole surveys for the final drillholes, along with the 5 holes drilled during the 2022 campaign (which could not be performed at the time due to poor ground condit ions), will be conducted during the next drilling campaign . The downhole survey was conducted by the drilling contractor using the REFLEX EZ -TRAC tool.

Metallurgical Drilling

Drilling and sampling activities were supervised by a suitably qualified Company geologist who was present at all times. All core 1 m drill samples were geologically logged by the geologist at the core warehouse.

The samples and composites produced should not be regarded as fully representative of the mineralisation as a whole, given the still limited understanding of the entire orebody. Nonetheless, it can be stated that no significant mineralogical variations are evident, although grade variations do occur, with samples from the northwest and south exhibiting higher grades than thos e from the northeast.

Aspects of the determination of mineralisation that are Material to the Public Report. In cases where ‘industry standard’ work has been done this would be relatively simple (eg ‘reverse circulation drilling was used to obtain 1 m samples from which 3 kg was pulverised to produce a 30 g charge for fire assay’). In other cases more explanation may be required, such as where there is coarse gold that has inherent sampling problems. Unusual commodities or mineralisation types (eg submarine nodules) may warrant disclosure of detailed information. RC Drilling RC drill samples are collected over 1 m intervals. The entire sample is placed in a properly named and numbered plastic bag. Samples are manually homogenised before being split using a three -tier riffle splitter to provide an approximate 3 –5 kg sample. Wet samples are split using a cone and qu arter method.

This phase of sample preparation is conducted in the field, and the resultant ~3 –5 kg samples are then transported to the Berkeley warehouse.

Samples are further split using a one tier riffle splitter such that 0.7 –1 kg samples are sent to the preparation laboratories of ALS (Seville, Spain) and analytical laboratory of ALS (Loughrea, Ireland). Samples are dried, fine crushed down to 70% below 2 mm, split to obtain 250 g and pulverised with at least 85% of the sample passing 75 µm. 10 g of sample is used for 51 elements analysis by ICP -MS method with results corrected for spectral inter -element interferences.

Metallurgical Drilling

DD drilling was used to obtain samples. The bulk metallurgical sample is a composite of selected samples from diamond drillholes.

Existing lithium and rubidium assay results from proximal RC drillholes were used to determine the 1 m intervals suitable to contribute to the selected intervals.

Drilling

techniques Drill type (eg core, reverse circulation, open -hole hammer, rotary air blast, auger, Bangka, sonic, etc) and details (eg core diameter, triple or standard tube, depth of diamond tails, face -

sampling bit or other type, whether core is oriented and if so, by what method, etc). RC Drilling Drilling comprised RC drilling using a 140 mm diameter face sampling hammer and DDH using PQ (85.40 mm core diameter), T101 (82.60 mm core diameter), and HQ (63.50 mm core diameter) diamond coring drill bits.

During PQ drilling, the wire -line system was utilised , and the core was oriented using the REFLEX ACT III tool in the first 31.40 m of the initial DD hole.

Subsequently, it was decided to use the conventional system using T101 drilling due to the challenging ground conditions in terms of hardness. This system did not allow for core orientation. HQ diameter

19 Criteria JORC Code explanation Commentary was only used in the first DD drillhole for the final 9.50 m.

Metallurgical Drilling

RC drillholes were drilled with a vertical dip. DDH are angled ranging from -40° to -70°. DD drilling was completed with PQ (85.4 mm core diameter), T101 (82.6 mm core diameter) and HQ (63.50 mm core diameter). Each 3 m of core sample are collected into se parate core boxes. PQ -diameter drilling was conducted using the wireline method, whereas T101 -diameter drilling was carried out using the conventional method. HQ diameter was only used in the first DD drillhole for the final 9.50 m.

The PQ core was successfully oriented using a Reflex ACT III tool.

However, due to technical constraints and the hardness of the material, the core diameter had to be slightly reduced to T101, at which point core orientation could no longer be achieved.

Drill sample

recovery Method of recording and assessing core and chip sample recoveries and results assessed. RC Drilling The RC drilling rig utilised suitably sized compressors to ensure dry samples were possible. Plastic sample bags are strapped to the cyclone to maximise sample recovery. Sample logs record whether the sample is dry, moist or wet.

When the drillholes pass through zones with a significant water inflow, the sample bags are pierced to release the amount of water from the cyclone. In this operation, a part of the finest fraction is lost, resulting in a poorer recovery. Future studies wi ll be undertaken to determine whether the loss of fines influences the grade. However, no material grade variation was observed across mineralised intervals that went from dry to wet in this drilling campaign.

Metallurgical Drilling

The DD drilling rigs used face discharge bits to ensure a low contact between the rock and drilling fluids, minimising ore washing. The length of each run is measured directly on the core and compared with the start and end depths recorded by the drillers. Recovery for each run is calculated in a spreadsheet, along with the overall core recovery for each DD drillhole, which were as follows:

• CCD007: 96% recovery • CCD021: 98% recovery • CCD028: 98% recovery The configuration of drilling and nature of materials encountered results in negligible sample loss or contamination. DD drillhole core was obtained in intervals between 0.10 m and 3 m.

Measures taken to maximise sample recovery and ensure representative nature of the samples. RC Drilling There is no current known relationship between sample recovery and grade. The RC and DD sample recoveries are of an acceptable level, and no bias is expected from any sample losses.

The Drilling Contractor possesses the requisite experience, and its drillers and assistants adhere to established industry protocols for drillhole operations.

The Company’s geologists supervised the drilling and were responsible for monitoring all aspects of the drilling, logging and sampling process.

Whether a relationship exists between sample recovery and grade and whether sample bias may have occurred due to preferential loss/gain of fine/coarse material. No bias related to preferential loss or gain of different materials occurred.

Logging Whether core and chip samples have been geologically and geotechnically logged to a level of detail to support appropriate Mineral Resource estimation, mining studies and metallurgical studies. RC Drilling Berkeley geological logging of RC chip samples included recording descriptions of lithology, weathering, alteration and mineralisation.

Metallurgical Drilling

Logging of DD core included recording descriptions of lithology, age, colour, oxidation, mineralisation, alteration, weathering, structures, textures, grain size and mineralogy.

20 Criteria JORC Code explanation Commentary

Geotechnical logging of DD drillhole core included recording descriptions of integrity (recovery and RQD), materials (lithology, rock strength and depth oxide staining), and structures (type, angle, contact type, infill, weathering).

All individual 1 m core intervals were geologically logged, recording relevant data to a set template using Company codes.

Whether logging is qualitative or quantitative in nature. Core (or costean, channel, etc .) photography. RC Drilling Geological logging is qualitative in nature.

Berkeley RC samples and chip trays are photographed.

Metallurgical Drilling

All DD core boxes were photographed. Logging is qualitative and records regolith, grain size, texture, mineralogy, lithology and geotechnical features. Core was digitally photographed.

The total length and percentage of the relevant intersections logged. 100% of RC drilled and DD core metres were geologically logged.

Sub-sampling

techniques If core, whether cut or sawn and whether quarter, half or all core taken. Metallurgical Drilling The purpose of core is to conduct metallurgical testing. 5 m intervals of each DD core were sent for metallurgical testing. They have been cut in half using a water saw, with care taken to ensure minimal ore loss. One half was placed into individually labe lled bags numbered by metre, while the other half will remain in the core boxes for future reference.

Core was cut using a diamond saw and quarter core submitted for metallurgical testwork.

and sample

preparation If non -core, whether riffled, tube sampled, rotary split, etc and whether sampled wet or dry. RC Drilling Berkeley RC drill samples were collected at 1 m intervals. RC intervals were sampled by splitting dry samples in the core shed to 3 – 5 kg using three -tier riffle splitter and further split to 0.7 –1 kg using a one-tier riffle splitter. Wet samples were split using a cone and quarter method.

For all sample types, the nature, quality and appropriateness of the sample preparation technique. RC Drilling RC drill samples were sent to ALS laboratory (Seville, Spain) for preparation and analysis. Samples were dried, fine crushed down to 70% below 2 mm and pulverised with at least 85% of the sample passing 75 µm. 10 g of sample was used for multi element anal ysis by ICP -MS method. This method is considered appropriate for this style of mineralisation.

Metallurgical Composite Sample DD core samples were sent to Oviedo University (Asturias, Spain) and SLR laboratory (former Wardell Armstrong International, Cornwall, United Kingdom) for preliminary metallurgical testwork including crushing, grinding, head sample characterisation, scanning electron microscope (SEM) mineralogical analysis, gravimetry, flotati on, and magnetic separation.

All DD drillhole core samples (15 m in total) were processed, with a combined weight of 47.4 kg.

Quality control procedures adopted for all sub -sampling stages to maximise representivity of samples. RC Drilling Routine sample preparation and analyses of RC samples were carried out by ALS, who operates an independent sample preparation laboratory in Seville (Spain) and an analytical laboratory in Loughrea (Ireland). QAQC procedures involve the use of duplicates wh ich are inserted into sample batches at a frequency of approximately 15%.

Metallurgical Drilling

The sample preparation techniques and QAQC protocols are considered appropriate for the nature of this testwork.

21 Criteria JORC Code explanation Commentary Measures taken to ensure that the sampling is representative of the in situ material collected, including for instance results for field duplicate/second -half sampling. RC Drilling Duplicate splits of RC samples are taken every 10 m downhole within the sampled intervals. The results from these duplicates generally show acceptable repeatability.

Metallurgical Drilling

Core was marked for sampling along an orientation line, and a consistent quarter of core was sampled along the DD drillhole. No field duplicates were completed for this program.

Whether sample sizes are appropriate to the grain size of the material being sampled. RC Drilling Sample sizes are well in excess of standard industry requirements.

Metallurgical Drilling

The sample size is considered appropriate for the nature of the testwork.

Quality of

assay data

and

laboratory

tests The nature, quality and appropriateness of the assaying and laboratory procedures used and whether the technique is considered partial or total. All samples were routinely assayed by ALS for 51 elements using ICP-MS method. This analytical method reports total content for each element.

Metallurgical Composite Sample The potential recovery of lithium, rubidium, and the other elements of economic interest, was tested using the following techniques:

• Gravity Release Analysis (GRA ) • Froth flotation • Magnetic separation For geophysical tools, spectrometers, handheld XRF instruments, etc, the parameters used in determining the analysis including instrument make and model, reading times, calibrations factors applied and their derivation, etc. No geophysical downhole tools were used.

Nature of quality control procedures adopted (eg standards, blanks, duplicates, external laboratory checks) and whether acceptable levels of accuracy (ie lack of bias) and precision have been established. RC Drilling Duplicate samples and blanks were regularly inserted into the sample stream by Berkeley, with approximately 13% of all samples related to quality control. The external laboratory used also maintain their own process of QAQC utilising standards, pulp repeat s, sample duplicates and blanks.

The review of the Berkeley quality control samples, along with the QAQC reports from the external laboratory, has indicated no issues with sample preparation, acceptable levels of analytical accuracy based on laboratory internal Standard Reference Materials (SRMs), and no bias present in the analytical datasets.

No project -specific external SRMs were inserted into the sample stream by the Company. This reflects the absence of commercially available standards considered representative of the Conchas mineralisation style and associated elemental suite. QAQC monitoring therefore relied on the use of blanks, field duplicates and ALS internal laboratory quality control procedures. The Competent Person (CP) recommends sourcing (or have prepared) and submitting external SRMs in future drilling campaigns, as a cri tical QAQC measure to establish confidence in the use of assay data.

Metallurgical Drilling

Acceptable levels of accuracy and precision were established in the preparation of the bulk sample composites.

Verification

of sampling

and assaying The verification of significant intersections by either independent or alternative company personnel.

No new drilling intersections are being reported.

The use of twinned holes. No twin holes were completed in either the 2022 or 2024 campaigns .

This was not warranted.

Documentation of primary data, data entry procedures, data verification, data storage (physical and electronic) protocols. RC Drilling All primary data was recorded in templates designed by Berkeley.

22 Criteria JORC Code explanation Commentary Assay data from the external laboratory is received in spreadsheets and downloaded directly into a Microsoft Access database managed by the Company. Data is entered into controlled Microsoft Excel templates for validation. The validated data is then loaded into a password secured relational database by a designated Company geologist. Daily backups of all digital data are undertaken. These procedures are documented in the Berkeley Technical Procedures and Protocols manual.

Discuss any adjustment to assay data. Lithium (ppm) assays received from the external laboratory are converted to Li 2O (ppm) using the stoichiometric factor of 2.153 and then passed to percentage.

Rubidium (ppm) assays are converted to Rb 2O (ppm) using the stoichiometric factor of 1.094 and then passed to percentage.

Tin (ppm) assays are converted to SnO 2 (ppm) using the stoichiometric factor of 1.270 and then passed to percentage.

Beryllium (ppm) assays are converted to BeO (ppm) using the stoichiometric factor of 2.775 and then passed to percentage.

Cesium (ppm) assays are converted to Cs 2O (ppm) using the stoichiometric factor of 1.060 and then passed to percentage.

Niobium (ppm) assays are converted to Nb 2O5 (ppm) using the stoichiometric factor of 1.431 and then passed to percentage.

Tantalum (ppm) assays are converted to Ta 2O5 (ppm) using the stoichiometric factor of 1.221 and then passed to percentage.

Location of

data points Accuracy and quality of surveys used to locate drill holes (collar and down -hole surveys), trenches, mine workings and other locations used in Mineral Resource estimation. RC Drilling The survey of drillhole collar locations has been completed by a qualified surveyor using standard differential GPS equipment achieving sub -decimetre accuracy in horizontal and vertical position.

Downhole surveys have been completed for 34 of the 36 holes drilled in the 2024 campaign. Downhole surveys for the final drillholes, along with the 5 holes drilled during the 2022 campaign (which could not be performed at the time due to poor ground condit ions), will be conducted during the next drilling campaign . The downhole survey was conducted by the drilling contractor using the REFLEX EZ -TRAC tool.

Metallurgical Drilling

A Leica iCON gps 70 series differential GPS was used to pick up the collars.

A REFLEX EZ -TRAC deviation tool was used for downhole surveying.

Specification of the grid system used. Universal Transverse Mercator (UTM) datum ETRS89 Zone 29 North.

Quality and adequacy of topographic control. Topographic control was based on a digital terrain model with sub -

metric accuracy sourced from the Spanish Geographical Institute (Instituto Geográfico Nacional) and was verified through detailed DD drillhole collar surveys by a qualified surveyor using a differential GPS.

Data spacing

and distribution Data spacing for reporting of Exploration Results. RC Drilling The objective of the 2024 drilling campaign was to check the results of previous exploration works which were based on soil sampling, geological mapping, and five RC holes drilled in 2022. The results reported in 2023 relate to three sections with a drillhole (CCR004) common to al l of them, with two or three holes on each section, with plan distances between drillholes from 130 m to 570 m.

In the 2024 campaign, the drillholes have been positioned around the four drillholes with the highest grades from the 2022 campaign, at distances of approximately 100 m from the central hole and 70 m between each hole. In some areas, the distance between d rillholes is as close as 25 m, allowing for the assessment of variability over short distances.

23 Criteria JORC Code explanation Commentary

Metallurgical Drilling

The DD drillholes were positioned with the objective of obtaining representative samples for metallurgical testwork from the three areas currently deemed most significant within the orebody.

The DD drillholes contributing to the metallurgical testwork program were drilled in three distinct locations within the orebody, spaced between 400 m and 800 m apart. Each DD drillhole may be considered representative of the specific zone in which it is s ituated, though not yet of the deposit as a whole.

Whether the data spacing and distribution is sufficient to establish the degree of geological and grade continuity appropriate for the Mineral Resource and Ore Reserve estimation procedure(s) and classifications applied. RC Drilling The data spacing is deemed sufficient to infer geological and grade continuity in the surveyed area, in support of Inferred Mineral Resources.

Whether sample compositing has been applied. RC Drilling No sample compositing was undertaken.

Metallurgical Composite Sample The sample was composited as described under Sampling Techniques in this Table 1.

Orientation of

data in

relation to

geological

structure Whether the orientation of sampling achieves unbiased sampling of possible structures and the extent to which this is known, considering the deposit type. RC Drilling The muscovitic leucogranite which hosts the mineralisation is ~2 km by ~1.2 km in area. The mineralised zone is interpreted to be sub -

horizontal following the regional plutonic tendency, with the potential for one or more feeder zones (i.e. based on geological logging, there are indications that drillholes in two areas have not reached the base of mineralis ation). No bias attributable to orientation of sampling was identified.

If the relationship between the drilling orientation and the orientation of key mineralised structures is considered to have introduced a sampling bias, this should be assessed and reported if material. RC Drilling All RC drillholes are vertical and perpendicular to the interpreted mineralised body. Due to the interpreted sub -horizontal nature of the mineralisation, no sampling bias is considered to have been introduced by the orientation of the drilling and the thic knesses are considered real.

Core orientation was only possible in drillhole CCD007 from 6 m to 31.40 m. Two fracture sets were identified. The dominant set is sub -

horizontal, while the secondary set dips at approximately 35° to the southwest. Other possible discontinuities, such as c ontacts or veins, were not encountered within the core orientation interval.

Metallurgical Drilling

All DD drillholes were planned and executed with an inclination in order to permit core orientation. Although the mineralisation is considered to be sub-horizontal, no bias attributable to orientation of drilling was identified.

Sample

security The measures taken to ensure sample security. RC Drilling Chain of custody is managed by Berkeley. Following initial splitting in the field, the resultant ~3 –5 kg subsamples were transported from the drill site by Company vehicle to a logging core shed where samples are prepared for dispatch. Samples are sent dir ectly from the core shed to the laboratory using a certified courier. Samples are included in cardboard boxes conveniently closed and strapped to prevent its opening. The samples are taken directly from the Berkeley facility to the external laboratory. Sam ple submission forms are sent in paper form with the samples, as well as electronically to the laboratory.

Reconciliation of samples occurs prior to commencement of sample preparation for assaying.

Metallurgical Drilling

Samples were stored in secure storage from the time of drilling, through cutting, bagging, and packing, and up until their dispatch to the laboratory.

24 Criteria JORC Code explanation Commentary

A reputable international transport company with shipment tracking enables a chain of custody to be maintained whilst the samples moved from Spain to the UK. Samples were again securely stored once they arrived and were processed at Baldhu, Cornwall.

At each point of the sample workflow the samples were inspected by a Company representative to monitor sample condition. Each laboratory confirms the integrity of the samples upon receipt.

Audits or

reviews The results of any audits or reviews of sampling techniques and data. No audits have been undertaken.

Section 2 Reporting of Exploration Results (Criteria listed in the preceding section also apply to this section.) Criteria JORC Code explanation Commentary

Mineral

tenement and

land tenure

status Type, reference name/number, location and ownership including agreements or material issues with third parties such as joint ventures, partnerships, overriding royalties, native title interests, historical sites, wilderness or national park and environment al settings. The Conchas Project lies on the Conchas I IP 6930 which is 100% owned by Berkeley Minera España, S.L., a wholly owned subsidiary of Berkeley Energia Limited under the General Regulations for the Mining Regime established under Royal Decree 2857/1978 of 25 August in Spain.

The Conchas I IP was originally granted in October 2020 for an initial three -year term. An extension of the Investigation Permit for a second three -year term (from October 2023) was granted in June 2024.

There are no historical sites, reserves or specially protected areas in the zone, which is primarily used for livestock grazing and agriculture.

The Conchas Project is located adjacent to the village of Fuentes de Oñoro and close to the border with Portugal.

The security of the tenure held at the time of reporting along with any known impediments to obtaining a licence to operate in the area. Tenure in the form of a granted IP and is considered secure. There are no known impediments to obtaining a licence to operate in this area.

Exploration

done by other parties Acknowledgment and appraisal of exploration by other parties. Mining in the area dates back to the World War II years when, in an artisanal manner, tin and tungsten were obtained by means of surface excavations and washed by hand.

Modern exploration at Conchas was carried out by Billiton PLC between 1981 and 1983. The investigation was focused on tin and tantalum, with lithium, rubidium etc. not taken into account . Billiton carried out several exploration work programs which resulted in a regional geological map and another detailed geological map, a leucogranite bottom isopach map, geochemistry with 85 test pits, trenches and 20 percussion drillholes, and sectiona l interpretations of the different magmatic facies.

SIEMCALSA (Mining Investigation and Exploration Society of Castilla y León, S.A.) within the European Union project POCTEP, summarised the Billiton data, making a review of the land and a chip sampling (14 samples) of the types of rocks existing in the area.

Mineralogical and metallogenetic studies of samples were carried out at the Universities of León (Spain) and Porto (Portugal), howe ver, Berkeley has not yet obtained access to these reports/results.

Only public domain historical data has been obtained by Berkeley.

Geology Deposit type, geological setting and style of mineralisation. Around the 70% of the permit area is filled by the Cenozoic cover and, only in the northwest, the Fuentes de Oñoro granite can be found.

Cenozoic materials have Oligocene age.

Granites make up the Vilar Formoso -Fuentes de Oñoro area, which in turn belongs to the Guarda Batholith whose origin is associated with the Hercynian orogeny. Regionally, coarse to very coarse -grained granodiorites and porphyritic granites are found, porphyritic and with a considerable amount of biotite, a rranged subparallel to the edge of the batholith and commonly considered as edge facies.

25 Criteria JORC Code explanation Commentary

The monzogranite facies is the one with the greatest superficial development and constitutes approximately 50% of the outcropping granites. They are two -mica granites, with a predominance of biotite, fine to coarse grain size and sometimes porphyry, although the potassium feldspar megacrystals do not reach the size of those of the previous edge f acies.

Aplogranites constitute the mineralised facies of aplo -pegmatitic leucogranites. This occurs in the vicinity of Fuentes de Oñoro and in front of the Portuguese town of Poço Velho. Preliminary mineralogy studies indicate the lithium, rubidium and caesium occurs in micas classified as intermediate between muscovite and zinnwaldite.

It also presents a millimetre mineralisation of cassiterite, and columbo -

tantalite distributed homogeneously throughout its surface. Cassiterite normally occurs in angular and heterometric crystals of between 10 μm and 1 mm. Tantalum and niobium occur in t he form of columbo -

tantalite, both in isolated crystals and in inclusions within the cassiterite.

Drill hole

Information A summary of all information material to the understanding of the exploration results including a tabulation of the following information for all Material drill holes:

o easting and northing of the drill hole collar o elevation or RL (Reduced Level – elevation above sea level in metres) of the drill hole collar o dip and azimuth of the hole o down hole length and interception depth o hole length. Exploration results are not the subject of this disclosure. All drill intercepts relating to the Conchas Project have been previously reported. The relevant stock exchange announcements included all collar and composite data and can be viewed on the Company’s website.

There are no further drillhole results that are considered material to the understanding of the exploration results. Identification of the broad zone of mineralisation is made via multiple intersections of drillholes and to list them all would not give the reader any further clarification of the distribution of mineralisation throughout the deposit.

If the exclusion of this information is justified on the basis that the information is not Material and this exclusion does not detract from the understanding of the report, the Competent Person should clearly explain why this is the case. Exploration results are not the subject of this disclosure. No information has been excluded.

Data

aggregation

methods In reporting Exploration Results, weighting averaging techniques, maximum and/or minimum grade truncations (eg cutting of high grades) and cut -off grades are usually Material and should be stated. Exploration results are not the subject of this disclosure. No data aggregation was required.

Where aggregate intercepts incorporate short lengths of high grade results and longer lengths of low grade results, the procedure used for such aggregation should be stated and some typical examples of such aggregations should be shown in detail. Exploration results are not the subject of this disclosure. No data aggregation was required.

The assumptions used for any reporting of metal equivalent values should be clearly stated. No metal equivalent values are reported.

Relationship

between

mineralisation

widths and

intercept

lengths These relationships are particularly important in the reporting of Exploration Results. If the geometry of the mineralisation with respect to the drill hole angle is known, its nature should be reported. The mineralisation is considered to be sub -horizontal. The inclination of the DD drillhole s was intended to facilitate core orientation and to verify any potential lateral variability in the mineralisation, which has not been observed in any of the drillholes.

If it is not known and only the down hole lengths are reported, there should be a clear statement to this effect (eg ‘down hole length, true width not known’). Exploration results are not the subject of this disclosure. No new drill intercepts are being reported.

Diagrams Appropriate maps and sections (with scales) and tabulations of intercepts should be included for any significant discovery being reported These should include, but not be limited to a plan view of drill hole collar locations and appropriate sectional views . Not applicable. Exploration results are not the subject of this disclosure.

The original exploration results and plan view of the drillholes for the samples used in relation to the metallurgical composite testwork are included in Berkeley’s announcement dated 29 January 2025 which is accessible on the Company and ASX websites.

Balanced

reporting Where comprehensive reporting of all Exploration Results is not practicable, representative reporting of both low and Exploration results are not the subject of this disclosure.

26 Criteria JORC Code explanation Commentary high grades and/or widths should be practiced to avoid misleading reporting of Exploration Results. All results are included in this report and in previous releases. These are accessible on the Company’s website.

Other

substantive

exploration

data Other exploration data, if meaningful and material, should be reported including (but not limited to): geological observations; geophysical survey results; geochemical survey results; bulk samples – size and method of treatment; metallurgical test results; bulk density, groundwater, geotechnical and rock characteristics;

potential deleterious or contaminating substances. Electrical Resistivity Tomography Survey Trial (July 2025) A trial of Electrical Resistivity Tomography (ERT) was conducted at the Conchas Project in July 2025. The trial survey was carried out by Análisis y Gestión del Subsuelo (AGS). The trial survey was supervised and coordinated by Berkeley. One of the trial’s aims was to delineate the boundaries of the various mineralised horizons.

The measurements were taken using a Syscal Pro 72 geophysical unit manufactured by IRIS Instruments (Orleans, France). It features an internal 250 W transmitter with a maximum output of 800 –1000 V (400 V was used in the trial) and 2.5 A, 10 simultaneous acquisition channels providing up to 1,000 r eadings per minute, and automatic switching for up to 72 electrodes.

The measurement procedure involved placing 72 electrodes along two profiles, each 710 m long oriented NE -SW and NW -SE, with a 10 m spacing between electrodes and penetration depth to 150 vertical metres. All electrodes were connected to the measurement equipment, and a specific sequential program was used to select which quadrupoles operated at each time and in what configuration.

For each profile, about 5,000 to 8,000 resistivity readings are collected, with each reading performed in stacks of 2 –4 repetitions. If successive readings deviate beyond tolerance, the instrument repeats the measurement up to 4 times, averages the result, and – if within tolerance – moves to the next reading. This process ensures accurate, reliable data for geophysical analysis.

GPS points were recorded with a Garmin eTrex 32x device at the start and end of each profile, as well as at the end of each cable segment along the profile, to precisely determine their location and to obtain the terrain topography, allowing reprocessing of data w.r.t surface relief.

Two measurement sequences were used:

• The dipole -dipole sequence, optimised for detecting horizontal resistivity changes, with 2,921 data points for high resolution.

• The Schlumberger reciprocal sequence, used for detecting vertical resistivity changes, with 2,195 data points, sensitive to vertical variations.

The forward modelling and the detection of two types of rocks with different resistivity in the surveyed area demonstrate the suitability and effectiveness of the ERT method for the study. This approach validates the identification of lithological contrast s and supports further interpretation of subsurface geological structures.

The final output is a terrain section displaying resistivity values in different colours, representing variations in this parameter.

The detected anomalies appear to correspond well with the two main types of rocks present in the study area (LGM and GMG) when compared with surface mapping and drillholes near the profiles.

The Company evaluated the data and determined that its quality was adequate for further analysis. The Company noted that there appeared to be sufficient contrast between different lithologies and that the results correlate well with the existing data, supporting the decision to undertake a broader geophysical campaign aimed at covering most of the deposit area.

Electrical Resistivity Tomography Survey (late 2025) Following the successful trial in July 2025 a comprehensive ERTS was completed in late 2025 comprising an additional 9 profiles ranging from 350m to 1,070 linear metres with 10m electrode spacing, for a total of 5,820 linear metres and 35,900 measurement p oints.

Maximum penetration depth was 153.3m.

Standard Operating Procedures and technical descriptions were identical to those described for the trial survey.

27 Criteria JORC Code explanation Commentary The survey was successful in providing additional resolution and modified scale and geometry of the host LGM unit, confirming (via reliable geophysical -lithological correlation) and extending possible LGM occurrences beyond previously known limits. This da ta has been incorporated into 3D modelling, refining mineralised unit boundaries.

Mineralogical Study

A check of the lithologies present in the zone, especially of the mineralised zone, in -situ analysis using a portable x -ray fluorescence (XRF) and two soil geochemistry campaigns (203 samples collected) were carried out.

Five RC drillholes were drilled in 2022. Several zones with significant mineralisation were intersected in four of these drillholes. Three mineralogical studies conducted by ALS, the University of Oviedo, and the University of Bilbao, on samples from drill holes and rock chip samples collected from outcrops within the mineralised zone have concluded that the mineralisation of Li, Rb, and Cs is associated with micas classified as intermediate between muscovite and zinnwaldite.

The Sn, Nb, and Ta mineralisation are more closely associated with the cassiterite present in the rock.

Density Determination Study In the three DD drillholes that were also drilled in 2024, 50 density measurements were taken across the different intersected lithologies, yielding the following values:

• All weathered leucogranite: 1.58 g/cm3 • Partially weathered leucogranite: 2.51 g/cm3 • Leucogranite: 2.64 g/cm3 • Regional granite: 2.64 g/cm3 Further work The nature and scale of planned further work (eg tests for lateral extensions or depth extensions or large -scale step -

out drilling).

Additional infill drilling is planned to improve confidence in the estimation of Mineral Resources and to satisfy geotechnical, metallurgical and hydrogeological data collection requirements as the project develops.

Diagrams clearly highlighting the areas of possible extensions, including the main geological interpretations and future drilling areas, provided this information is not commercially sensitive. Refer to diagrams and plan views disclosed in previous announcements. These are accessible on the Company’s website as discussed above.

Section 3 Estimation and Reporting of Mineral Resources (Criteria listed in section 1, and where relevant in section 2, also apply to this section.) Criteria JORC Code explanation Commentary

Database

integrity Measures taken to ensure that data has not been corrupted by, for example, transcription or keying errors, between its initial collection and its use for Mineral Resource estimation purposes. Site visit review by Competent Person (CP) included database checks against field, hand logged data and assay certificates and drill collars.

Data validation procedures used. Extraction from the site supplied data to Datamine software for Mineral Resource estimation included both logical and visual checks of both the drillhole and assay data. Items included checking for under/overlaps, missing data and statistical analysis of grade

distributions

Site visits Comment on any site visits undertaken by the Competent Person and the outcome of those visits. A site visit was completed by CP Galen White of Maja Mining Limited during January 2026. His review included drillhole locations, geology logging and interpretation, assay validation and QAQC review and a general overview of all current and previous exploration. It is the opinion of the CP’s that all relevant data was appropriate for completion of a Mineral Resource estimate (MRE).

If no site visits have been undertaken indicate why this is the case. N/A. Site visit was undertaken .

Geological

interpretation Confidence in (or conversely, the uncertainty of) the geological interpretation of the mineral deposit. The minerals and metals of interest are hosted in a muscovite leucogranite ( LGM), additional low -grade mineralisation is associated with a surface altered igneous basement (CSG) and

28 Criteria JORC Code explanation Commentary proximal regional granite (GMG).

Nature of the data used and of any assumptions made. RC sampling was used to determine both geological and grade domain boundaries.

The effect, if any, of alternative interpretations on Mineral Resource estimation. Due to the relatively simple geology – all high-grade ( HG) mineralisation is confined to the LGM unit, which is relatively homogeneous – an alternative interpretation is improbable.

The use of geology in guiding and controlling Mineral Resource estimation. HG mineralisation is constrained by the LGM geological unit. Grade continuity is affected by mica intensity which is host to the dominant metals of interest (rubidium, lithium, caesium, beryllium), with associated heavy minerals containing tin, tantalum an d niobium.

The factors affecting continuity both of grade and geology. HG mineralisation is constrained by the LGM geological unit. Grade continuity is affected by mica intensity which is host to the dominant metals of interest (rubidium, lithium, caesium, beryllium), with associated heavy minerals containing tin, tantalum an d niobium .

Dimensions The extent and variability of the Mineral Resource expressed as length (along strike or otherwise), plan width, and depth below surface to the upper and lower limits of the Mineral Resource. The current MRE is divided into a north and south LGM zone, primarily controlled by drill access. The northern zone is roughly 800 m EW x 400 m NS to a maximum depth of 180 m. The southern zone is roughly 450 m EW x 350 m NS with an average depth from 40 m to 50 m deep. The mineralisation extends to the surface CSG with a low -grade cap which is 0 –5 m thick. Proximal low -grade mineralisation extends into the GMG with some internal interbeds in the LGM.

Estimation and

modelling

techniques The nature and appropriateness of the estimation technique(s) applied and key assumptions, including treatment of extreme grade values, domaining, interpolation parameters and maximum distance of extrapolation from data points. If a computer assisted estim ation method was chosen include a description of computer software and parameters used. Grade estimation of HG mineralisation was completed using ordinary kriging (OK) , based on subdivision into appropriate grade domains containing near log normal grade distributions. Variograms were modelled and top capping was applied where appropriate to reduce the local influence of isolated HG outliers. Samples were generally 1 m in length. Any samples not 1 m were composited to 1 m. Grades were estimated into 40 m x 40 m x 5 m size panels.

Geology boundaries were honoured using 5 m x 5 m x 1 m sub -

cells. HG mineralisation was constrained by a hard boundary based on interpretation using the drillhole sample data. Peripheral drillholes were given an influence of up to 80 m. Low grade and waste material grades were estimated using inverse distance squared. Datam ine software was utilised for domaining, block model construction and grade estimation.

Drillhole and Sample Statistics and Variogram Modelling :

Population statistics were reviewed for all grade estimation domains, including top capping requirements which were required as spatially isolated high -grade outliers are present.

Mineralisation is hosted in two product groups – Mica, containing Rb, Cs, Li and Be; and Heavy Minerals (HM) made up of Sn, Ta and Nb. Both groups generally fall within the same common spatial mineralisation domain.

The Mica group currently has the greatest economic value, so the Mica domain has been used to define the main zones (North and South) of economic interest for both groups.

Rb has the best -behaved lognormal distribution within the Mica group with a very low coefficient of variation. Both Li and Cs show a good correlation with Rb, with Be having a very low correlation with Rb.

Rb was used for the variography modelling for the Mica group, with Li and Cs being linked to Rb for grade estimation to maintain the grade correlation between the three major Mica elements. Be was treated independently.

Nb was used for the variography modelling for the HM group, with Sn and Ta being linked to Nb for grade estimation.

Variograms were modelled for Rb, Be and Nb to determine spatial continuity parameters for grade estimation using OK.

29 Criteria JORC Code explanation Commentary Nugget effect for both Mica domains is around 15 –20% which indicates that the area closest to each sample will be strongly influenced by the local grade, i.e. less smoothing required and reasonable reliability for local grade estimates. The nugget for the HM domains at around 5% demonstrates very good confidence in local grade estimates.

In the Mica group, around 60% of the deposit variability occurs with a 60–100 m range, demonstrating that a drill pattern of around 40 m spacing will produce a high confidence grade estimate. Note that the grade continuity of the HM group elements is slightly higher than the Mica group elements.

Generally, the Northern domain has slightly better continuity than the Southern domain.

The shape and nature of the variograms combined with the reasonably well -behaved sample grade populations support the use of OK for grade estimation.

The availability of check estimates, previous estimates and/or mine production records and whether the Mineral Resource estimate takes appropriate account of such data. This is the maiden MRE. No previous production has taken place.

No alternative estimation methods were used.

The assumptions made regarding recovery of by -products. No recovery of by -products has been included. All potential products are included in the MRE statement.

Estimation of deleterious elements or other non -grade variables of economic significance (eg sulphur for acid mine drainage characterisation). No deleterious elements have been identified that are considered likely to materially affect the potential extraction or processing of the mineralisation. The deposit is hosted within a muscovitic leucogranite and does not contain significant sulphide mineralisation that would typically give rise to acid mine drainage concerns. No deleterious elements or other non -grade variables of economic significance have been estimated in the current MRE.

Further metallurgical and environmental studies will be undertaken as part of future project evaluation.

In the case of block model interpolation, the block size in relation to the average sample spacing and the search employed. The average drill spacing ranges from 50 m to 80 m. An estimation panel size of 40 m x 40 m x 5 m was selected as being appropriate based on the drill spacing.

Any assumptions behind modelling of selective mining units. No selective mining has been considered. The HG zone can be bulk mined without any significant dilution or ore loss.

Any assumptions about correlation between variables. Rubidium, caesium and lithium all correlate with each other.

Estimation parameters used for rubidium were applied to the other two elements. Beryllium was treated separately. Tin, tantalum and niobium correlate with each other. Estimation parameters used f or tantalum were applied to the other two elements.

Description of how the geological interpretation was used to control the resource estimates. Hard mineralisation boundaries were interpreted based on a net smelter return (NSR ) value for the higher -grade mineralisation. Low grade and waste were separated from the higher -grade mineralisation to avoid any grade smoothing issues.

Discussion of basis for using or not using grade cutting or capping. Top capping was used.

The process of validation, the checking process used, the comparison of model data to drill hole data, and use of reconciliation data if available. The model was validated visually, statistically and through the use of swath plots.

Moisture Whether the tonnages are estimated on a dry basis or with natural moisture, and the method of determination of the moisture content. Tonnages are estimated on a dry basis. Moisture content was not estimated. A dry bulk density of 2.64 g/cm³ was adopted based on 50 direct density determinations completed on representative PQ DD drill core samples using the Archimedes immersion method.

Cut-off

parameters The basis of the adopted cut -off grade(s) or quality parameters applied. At current market revenues the calculated breakeven NSR is US$18/t. However; Conchas does have two significant HG zones with good mining continuity at the US$100/t NSR cut -off. To reduce the risk of potential market over supply and any future decrease in product revenue, reporting the MRE at the higher NSR cut -off of US$100/t is considered prudent.

Sources used for determination of saleable revenue in Euros per tonne are summarised below.

30 Criteria JORC Code explanation Commentary

Mining factors

or assumptions Assumptions made regarding possible mining methods, minimum mining dimensions and internal (or, if applicable, external) mining dilution. It is always necessary as part of the process of determining reasonable prospects for eventual economic extraction to consider potential mining methods, but the assumptions made regarding mining methods and parameters when estimating Mineral Resources may not always be rigorous. Where this is the case, this should be reported with an explanation of the basis of the mining assumptions made. Open pit mining was selected due to the mineralisation being exposed or close to the surface, with generally horizontal mineralisation geometry. Bulk mining with little to no internal waste, dilution or ore loss has been considered. Reasonable prospects fo r eventual economic extraction (RPEEE) has been employed though Whittle open pit optimisation using appropriate costs, recovery, payability and metal prices. The MRE was constrained by the Whittle open pit shell.

Parameters used for the Whittle open pit optimisation runs are summarised below .

RPEEE – Cost Assumptions are summarised below .

Metallurgical

factors or

assumptions The basis for assumptions or predictions regarding metallurgical amenability. It is always necessary as part of the process of determining reasonable prospects for eventual economic extraction to consider potential metallurgical methods, but the assumption s regarding metallurgical treatment processes and parameters made when reporting Mineral Resources may not always be rigorous. Where this is the case, this should be reported with an explanation of the basis of the metallurgical assumptions made. Metallurgical parameters were obtained for testwork completed and applied in the RPEEE parameters.

Environmental

factors or

assumptions Assumptions made regarding possible waste and process residue disposal options. It is always necessary as part of the process of determining reasonable prospects for No assumptions have been made.

31 Criteria JORC Code explanation Commentary eventual economic extraction to consider the potential environmental impacts of the mining and processing operation. While at this stage the determination of potential environmental impacts, particularly for a greenfields project, may not always be well ad vanced, the status of early consideration of these potential environmental impacts should be reported. Where these aspects have not been considered this should be reported with an explanation of the environmental assumptions made.

Bulk density Whether assumed or determined. If assumed, the basis for the assumptions. If determined, the method used, whether wet or dry, the frequency of the measurements, the nature, size and representativeness of the samples. Bulk density was determined from direct measurements completed on representative PQ DD drill core samples collected from drillholes CCD007, CCD021 and CCD028. A total of 50 dry bulk density determinations were completed, comprising 16 measurements from CCD 007, 17 measurements from CCD021, and 17 measurements from CCD028.

Density measurements were undertaken using the Archimedes immersion method on dry core samples. The dataset included measurements from muscovitic leucogranite, altered muscovitic leucogranite, regional muscovitic granite and weathered material.

Three control measurements were completed during each measurement session to verify weighing system stability, resulting in a total of nine control measurements.

Based on the measured density dataset, a dry bulk density of 2.64 g/cm³ was adopted for the MRE.

The bulk density for bulk material must have been measured by methods that adequately account for void spaces (vugs , porosity, etc), moisture and differences between rock and alteration zones within the deposit. Density measurements were completed using the Archimedes immersion method, which is considered appropriate for determining the bulk density of competent crystalline rocks and inherently accounts for sample volume, porosity and void space characteristics.

Measurements were completed on the principal lithological and weathering units present within the deposit. Average densities of 2.64 g/cm³ and 2.64 g/cm³ were obtained for muscovitic leucogranite and regional muscovitic granite respectively, while altered muscovitic leucogranite returned an average density of 2.53 g/cm³ and weathered material returned an average density of 1.58 g/cm³.

Discuss assumptions for bulk density estimates used in the evaluation process of the different materials. The mineralised domains are hosted predominantly within muscovitic leucogranite and regional muscovitic granite, both of which display very similar density characteristics. Consequently, a density of 2.64 g/cm³ was considered representative of the minerali sed material included in the MRE. Weathered material represents only a minor proportion of the model and is not considered material to the global resource estimate.

Classification The basis for the classification of the Mineral Resources into varying confidence categories. All of the MRE is currently classified as Inferred, until additional drilling and test work is completed.

Whether appropriate account has been taken of all relevant factors (ie relative confidence in tonnage/grade estimations, reliability of input data, confidence in continuity of geology and metal values, quality, quantity and distribution of the data). All current factors have been considered in the classification.

Whether the result appropriately reflects the Competent Person’s view of the deposit. The CP is comfortable with the Inferred classification for the Conchas deposit.

Audits or

reviews The results of any audits or reviews of Mineral Resource estimates. No audits or third -party reviews have been completed.

Discussion of

relative

accuracy/

confidence Where appropriate a statement of the relative accuracy and confidence level in the Mineral Resource estimate using an approach or procedure deemed appropriate by the Competent Person. For example, the application of statistical or geostatistical procedures to quantify the relative accuracy of the resource within stated confidence limits, or, if such an approach is not deemed appropriate, a qualitative discussion of the factors that could affect the relative accuracy and confidence of the estimate. The MRE is accurate within the definition of an Inferred MRE. In the CP’s opinion that implies a level of accuracy of ±30% to 50%.

32 Criteria JORC Code explanation Commentary The statement should specify whether it relates to global or local estimates, and, if local, state the relevant tonnages, which should be relevant to technical and economic evaluation. Documentation should include assumptions made and the procedures used. The estimate should be considered global.

These statements of relative accuracy and confidence of the estimate should be compared with production data, where available. No production data is available.