Layer adds MiniMax H3 for stereo 2K text, image, and reference-to-video generation · cho.sh
Layer adds MiniMax H3 for stereo 2K text, image, and reference-to-video generation
AI video model releases
Layer adds MiniMax H3 for stereo 2K text, image, and reference-to-video generation
MiniMax H3 is now available in Layer for text-to-video, image-to-video, and reference-to-video generation at stereo 2K resolution; Layer says it has strong motion consistency across all three modes.
Layer added six Tripo mesh operations: Rig, Animate, Retarget, Segment, Remesh, and Retexture, plus a multi-part mesh viewer for 3D asset production inside the platform.
The Layer agent can run parallel sub-agent streams within one chat turn, allowing multiple sub-tasks to execute concurrently instead of sequentially.
A bug fix restores selection of a specific video or mesh frame as a reference image input, while chat streams now recover from interruptions mid-generation.
Workspace-level AI controls now automatically propagate to sub-workspaces, and disabled models no longer appear in workspace model pickers.
cho.sh