Dreamina Introduces Seedance 2.5 To Scale Script-Based Commercial Production Efforts
New York, United States, July 7th, 2026, FinanceWire
Dreamina announced the launch of Seedance 2.5, a new model integrated into the company’s AI video tools that reframes commercial production around script-driven workflows and supports direct generation of continuous short-form video sequences.
Positioned as a suite of competencies for building out commercials from concept rather than solely on-location footage, Seedance 2.5 allows creators to write a script that can double as a shoot brief. Encoding camera moves, style, tone, pacing, and action into a textual format allows the model to interpret a wider array of production directions and serve them via generated video.
Expanding on existing updates to Dreamina’s video generators, Seedance 2.5 supports native generation of 30-second video clips. Rather than chaining together disparate short clips to construct a continuous product video or commercial, Seedance’s video output manages pacing, shot changes, and emotional tone over an entire 30-second interval. As such, users can write scripts that act as self-contained videos, complete with instructions on shot length.
Supporting multimedia input allows for more expansive scripts that draw from a variety of influences. In addition to text prompts, Seedance 2.5 accepts storyboards, product images, reference photos, mood music or soundtrack MP3s, camera-lens directions, character design references, and scene notes. Images and multimedia uploaded to the model can be referenced by the submitted script, allowing users to link dialogue or scene directions to visual and audio cues that Seedance 2.5 can interpret when filming digitally.
The Seedance 2.5 model can natively edit portions of generated videos. Supplying the model with an output clip allows users to direct Seedance to re-shoot certain portions of the video without re-rendering static frames. Customers who wish to adjust camera moves, pacing, or replace elements of a scene can indicate which sections should be replaced, and have Seedance re-film those passages while leaving the rest of the video intact. This editing workflow is presented as part of Seedance’s pipeline for running iterations on videos produced through its script-focused approach.
Seedance 2.5 comes at a time when marketers are rethinking the limits of commercial video. The model will be marketed as a way to create multiple assets from a single production effort: brands can use a single script to produce a long-form, cinematic version for landing page use or digital OOH, a quicker-paced social-ready edit built for video apps and sites, an un-paced demo version, and niche edits for holidays or targeted audiences. One example shared by Dreamina illustrates how a skincare brand could produce a master script for an everyday moisturizer that feeds TV, social, holiday, and demo versions of a commercial without requiring additional shoots or expensive re-shoots.
Similarly, a consumer electronics brand could generate a single script for a new wearable device that produces a cinematic hero video for its landing page, a fast-cut teaser for social platforms, a feature-focused explainer for retail partners, and a limited-time promotional cut for seasonal sales campaigns.
Dreamina’s video generation model furthers its emphasis on scripts as central planning documents. Where past video tools may have accepted scripts as written dialogue and simple location-direction, Seedance 2.5 allows users to describe camera movements (“pan right to reveal moisturizer”), lighting (“increase brightness to match her radiant skin”), object motion (“have the bottle float onto her desk”), and target emotional response (“viewer should feel thirsty”) within their scripts. Details included in a Seedance-compatible script are interpreted by the model during generation and used to construct video scenes, placing more emphasis on writers and treatment developers as producers over directors or on-set managers.
Building on past work with multimodal video generation, Seedance 2.5 integrates with Dreamina’s existing production tools. Compatible with Dreamina, inputs to the Seedance model can be crafted across media types before prompting video generation. Reference materials, audiovisual inspiration, and direction can be appended to scripts within Dreamina to feed Seedance 2.5 during production. Dreamina framed the announcement as a tool for brands and marketers seeking greater control over the narrative development phase of ad creation.
Highlights from today’s release include interpreting scripts as shoot-ready documents, continuous video generated from prompt-embedded direction, multimodal inputs simplify reference material collection, and patch editing eliminates generation bottlenecks. Dreamina closed the announcement with an overview of Seedance 2.5’s intention to enable story-driven production: When writers and marketers focus on core narratives, Shot cuts and camera directions can be translated to programmatic instructions.
About Dreamina
Dreamina specializes in AI video generation and editing tools for commercial video production. Dreamina constructs and tunes media-focused generative models and develops video production tools that accommodate multimedia inputs and embed narrative planning into the content creation process. Dreamina serves businesses, marketing teams, and creators who are building short-form content and commercial assets.
Dreamina is part of CapCut's broader ecosystem of content creation tools, giving brands a connected pipeline from scriptwriting through final video export. The platform is increasingly used across e-commerce, beauty, and consumer tech sectors to accelerate campaign turnaround without sacrificing production value or creative control.
Website: https://dreamina.capcut.com/
Contact
Dreaminacontact@dreamina.capcut.com
Disclaimer. This is a paid press release.