New AI models, more video input options, and smoother team work on the canvas.
ADDED
-
Hailuo MiniMax H3 Max — use H3 for fast iterations and H3 Max for final generation.
-
Claude — added Fable 5, Sonnet 5, and Opus 5.
-
Wan Video v3 — use documents, images, video, and audio as inputs.
-
Media player — control video and audio on the canvas from one window.
-
Profile avatars — upload a photo or pick a color so teammates are easier to tell apart.
UPDATED
-
Gemini Omni 1.1 — now supports quality up to 4K.
-
Video sockets — @mentions now work for video, and nodes inside frames snap to the frame edges.
Open your workspace to try the latest improvements.