Omni API Docs
Overview
Build with Omni API model groups, tasks, credits, and webhooks.
Omni API documentation
Omni API exposes production image, video, and multimodal model groups through one task API. Use the model catalog to choose a model group, upload input files through signed URLs, submit a task, and poll or open task details from Console.
Core flow
- Create an API key in Console.
- Fetch the public model catalog from
GET /api/omni/models. - Upload input files with signed upload URLs when the model requires files.
- Create a task with
POST /api/omni/tasks/create. - Poll
GET /api/omni/tasks/{portal_task_id}or use status batch endpoints. - Download outputs from the returned manifest when the task succeeds.
Ready model groups
| Model group | Category | Status |
|---|---|---|
| Omni Image AIO | Image | Ready |
| Omni Video AIO | Video | Ready |
| Omni Multi AIO | Multi | Ready |
Omni Multi AIO includes image upscaling, audio transcription, voice-clone TTS, VLM chat, and Talk to Prompt on a shared runtime that maintains multiple model groups.
Console
Console contains task history, task details, credits, billing, payments, usage, and API keys. Anonymous visitors can browse model and pricing pages, but task submission and API usage require a signed-in account.