Dreamina launches short-drama studio built on DreamScript, Anycook and DreamShot models
ByteDance's generative media app Dreamina announced on October 26 that its short-drama studio is live. It combines three in-house models, DreamScript, Anycook and DreamShot, with post-production tools and a remake mode for existing dramas.
Dreamina, ByteDance's generative media app, announced on October 26 that its short-drama studio is now live. The product packages three in-house models into a single production workflow for the vertical-video drama format, according to Pandaily.
The three models
DreamScript (script model). It generates scripts across the formats that dominate the genre: original stories, adaptations, overseas versions and commentary-style cuts. It also offers what Dreamina calls hit diagnosis, which compares a draft against platform trends and suggests structural changes. It can read and break down texts of up to 100,000 characters.
Anycook (image model). It is tuned for short-drama art direction. Dreamina says it is calibrated to produce characters that stay visually distinct and consistent across a production, and to build scenes in a range of visual styles to suit different story settings.
DreamShot (storyboard model). It works in two modes, a free canvas and a structured storyboard. It supports global edits, so a change made through dialogue with the tool propagates across the board rather than applying one frame at a time. A 3D director's table lets a creator place a virtual camera, set its movement and preview the resulting shot before anything is rendered. The system can also string key shots together into a rough animated pass to check pacing and continuity.
Post-production layer
The studio adds several editing tools around the models:
- In-place reshoots: individual shots can be reshot without regenerating the whole piece.
- Audio separation: dialogue is split from background sound so a scene can be redubbed.
- Shot splitting: a long sequence is cut automatically into editable units.
- Asset carry-over: assets created elsewhere in Dreamina move into the studio, with edits applying along the chain.
Remake mode
The feature most clearly aimed at distribution is remake mode. A user uploads an existing short drama, and the system identifies its characters, sets, props and shot structure. It then regenerates the visuals in a chosen style, aspect ratio and target region while keeping the original pacing and camera language. Pandaily describes this as a localisation pipeline for a format that Chinese studios have been exporting aggressively.
What is not yet disclosed
Pricing, parameter counts, context windows other than the 100,000-character script input, and benchmark scores for the three models have not yet been disclosed.
What this means
Pandaily reports that ByteDance has approached the AI short-drama market from several directions, from video models to platform incentives for AI-made dramas, and that this is the first time it has offered the whole chain in a single product. Script, image and storyboard models now sit in one workflow with shared assets and edits that propagate along the chain.
The design choices point to the workflow problems Dreamina is targeting: character consistency across a production, previewing camera moves before rendering, and reshooting single shots without regenerating a whole piece. Remake mode extends the studio to existing dramas, since it regenerates visuals for a chosen style, aspect ratio and target region while retaining the original pacing and camera language.
Related Articles
ByteDance's Seedance 2.0 Pro tops HKU video AI ranking with 3.84 out of 5 for visual quality
A University of Hong Kong study ranked 12 leading AI video-generation models and found Chinese models ahead of US rivals on visual quality. ByteDance's Seedance 2.0 Pro scored 3.84 out of 5, while several Chinese models fall behind on safety safeguards, according to the researchers.
GitHub Copilot assisted permissions now generally available in CLI, app and VS Code
GitHub announced that assisted permissions are now generally available in GitHub Copilot CLI, the GitHub Copilot app, and Visual Studio Code. The feature applies when Copilot requests permission to run a command or access a file system path.
OpenAI GPT-6.1 Terra Pro lists on OpenRouter at $1 input, $5 output per 1M tokens, 1.1M context
GPT-6.1 Terra Pro is the same underlying model as GPT-6.1 Terra, served with reasoning.mode set to pro. OpenRouter lists it with a 1.1M context window at $1 input and $5 output per 1M tokens, and says a typical request costs several times more than on Terra.
GPT-6.1 Terra is now generally available in GitHub Copilot
GitHub announced that GPT-6.1 Terra, which it describes as OpenAI's newest model, is now generally available in GitHub Copilot. The model is designed for everyday agentic coding and terminal work at a lower cost than GPT-6.1 Sol.
Comments
Loading...