Video Generation Is Live: 16 Models in Your Chats
As of July 2026, video generation is live across idapt: 16 models including Google's Veo family, Runway, Kling, Luma, Seedance, and Hailuo, all in the same composer as your chats.
How it works
Describe the shot in any chat, or attach an image to animate as the first frame. Generation runs asynchronously (you keep working; the clip arrives), and every clip lands in a Generated Videos folder in your Drive: shareable, referenceable in future chats, yours.
- Text-to-video and image-to-video, per model.
- Native audio on the models that support it (the Veo family, Kling, Seedance, Grok Imagine): dialogue, effects, ambience.
- Per-second billing at the model's rate, shown before you generate; failed generations refund automatically.
- Resolutions up to 4K on the flagship lanes.
Picking a model
The spread is real: a quick draft on an affordable model costs a few cents; a flagship 4K shot costs a few dollars. The pattern that works is drafting cheap and finishing premium, and because every model shares one composer, switching is a dropdown, not a different product. The video models directory lists live specs and prices, and the best video generators roundup compares the leaders by job.
Good to know
- Video bills against paid plans or Usage Credits; there is no free video lane.
- Clip length and aspect options vary per model and are shown in the composer.
- Agents can generate video too, subject to the same budgets as any tool.
Get started
Open a chat, switch the composer to video, and describe a shot. The feature page covers formats and limits, and the side-by-side comparison surface lets you run one brief across two models before you commit.
Found this helpful? Share it: