About the Models

About the models

We currently use our Gen 7.2 model, and unlike others, we don't resell model access. All our models originate from the open-source community or our own R&D — we then disassemble, modify, fine-tune and optimize them to align with our legacy and 2DAI❤️ART lines. They also run on our own private cloud network.

Gen8 Preview video — the first public release of our next video engine is live behind generate_video with videoModel: "next": clips with sound, from a first frame plus up to 6 references that stay consistent, and in the studio an audio reference (music, or a voice to clone) with TIXI writing the scenario. fast quality and 1 to 10 s from Holder; every preset and clips up to 20 s at Founder. get_account says whether it is open to the account (videoDurationsNext present) and what it may submit.

Gen8 Preview images — the image side of the same generation is live behind generate_image with imageModel: "next": up to 6 references kept natively, legible text in the picture, long literal prompts honoured, the same quality ids (Fast to Max from Holder, Ultra from Believer, Ultimate from Supporter), priced on render time, roughly four to seven times the Gen 7.2 preset; get_account reports canUseImageNext and qualities.imageNext. Creations are stamped aiModel: "gen8.0" with imageModel: "next".

What that means in practice

• Same weights everywhere. The images and videos you generate through this MCP server come off the exact same Gen 7.2 and Gen8 stack as the ones you get from the 2DAI studio at 2dai.io — no watered-down "developer tier", no throttled model.
• In-house LLM rewriter. Prompt enhancement (the "TIXI" pass that expands short prompts) runs on a proprietary rewriter we operate — not a re-sold third-party LLM.
• Private cloud. Both the image/video generator and the rewriter run on network infrastructure we own end-to-end. Your prompts and outputs never transit through a public multi-tenant model API.

Runtime signal for agents

Every get_account response carries a platform block:

{
  "platform": {
    "model": "Gen 7.2",
    "privateCloud": true,
    "about": "We currently use our Gen 7.2 model, …"
  }
}

Server-level MCP instructions also mention the stack, so an agent that reads its briefing at connect time can answer "what powers you?" without an extra tool call.

Versioning

The model version bumps forward when we retrain — get_account.platform.model is always authoritative. The MCP server itself follows its own semver independently of the model version.