Appearance
Vercel AI
Vercel AI Gateway (/media/gateway) routes image and video generation through Vercel's unified gateway, giving you access to hundreds of models behind a single API key with automatic provider failover.
Where to configure
Configure the Vercel AI Gateway provider once under Settings → AI. Gateway is a universal-credential provider, so the same API key drives both LLM and media generation; no separate Settings → Media entry is needed. See Settings for provider setup.
Tabs / operations
| Tab | Operation | Key fields | Produces |
|---|---|---|---|
| Text → Image | Image | Model (discovered live), prompt, size, number of images | Image(s) in /files |
| Text → Video | Video | Model (discovered live), prompt, duration, aspect ratio | MP4 clip in /files |
| Image → Video | Video | Source image, model, prompt, duration | MP4 clip in /files |
Generation flow
Images are delegated to the matching AI-SDK provider and complete synchronously. Video uses AI SDK v6's experimental generateVideo and is also completed inline, with the Undici timeout extended to 15 minutes to accommodate long renders. Finished artifacts land in the Media Library and appear in the Render Queue with Edit in Designer and Post hand-offs. See Media Studios overview for the shared flow.
Caveats
- Models are discovered live from the Gateway catalog (
/v1/models) and filtered by modality. If the catalog call fails, the dropdown falls back to the descriptor's defaults plus free entry. - Video is synchronous and long-running; large renders are rejected before base64 inflation if they would exceed the 512 MB artifact limit.
- Audio/speech is not exposed by the Gateway AI-SDK provider, so the studio offers image and video only.
Related docs
- Media Studios overview — shared render-queue and hand-off flow.
- Media Library — where finished images and video are saved.
- Settings — configuring AI providers.
Verified against v1.0.0