Ant Group's inclusionAI released open weights for Invoker-Render 1.0, a 33B model that turns a rough mock-up of a scene into realistic video 🎬 Think of it as a game engine plus an AI renderer: • You give it a simple stand-in for the scene (colored layout maps, 3D boxes, shapes or stick-figure skeletons) and the camera moves you want • Plus one frame that shows how it should look, and a text prompt • It generates the finished video, following your camera path and how things move • It streams, so the video keeps going as long as new scene input comes in Built on MiniMax's open H3 video model. It's big: ~132 GB, set up to run on 8 GPUs. Non-commercial license. Paper coming soon. https://huggingface.co/inclusionAI/invoker-render-1.0
New model detected · · aitracker.bot
invoker-render-1.0
by inclusionAI found in inclusionAI on Hugging Face (hand post, owner OK)
New model
- Status
- Released
- Maker
- inclusionAI
- Found in
- inclusionAI on Hugging Face (hand post, owner OK) huggingface.co
- Detected
- Posted to X
- 0 s after detection
- Model ID
invoker-render-1.0also listed asinclusionAI/invoker-render-1.0
Where to use it
first seen Nov 17, 2026| Where | Listed | After first sighting |
|---|---|---|
| Hugging Face weights | first |
Timeline
1 sighting-
release
Spotted in inclusionAI on Hugging Face (hand post, owner OK)
Ant Group's inclusionAI released open weights for Invoker-Render 1.0, a 33B model that turns a rough mock-up of a scene into realistic video 🎬
Post on X