> ## Documentation Index
> Fetch the complete documentation index at: https://docs.omniall.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Runway

> Runway video via unified task API and official-style paths.

# Runway

Call **Runway** on Omniall AI in two styles:

| Style | Create | Poll |
| - | - | - |
| **Unified (recommended)** | `POST /v1/task/submit` | `GET /v1/task/{task_id}` |
| **Official** | `POST /runwayml/v1/image_to_video` · `POST /runwayml/v1/video_to_video` | `GET /runwayml/v1/tasks/{task_id}` |

Unified aliases: `POST|GET /runway/v1/task/...`.

**Host:** `https://api.omniall.ai` · Auth `Authorization: Bearer sk-...`.

## Models

Examples: `gen4_turbo`, `gen4_turbo/image-to-video`, `gen3a_turbo`, `gen4_aleph`, `gen4_aleph/video-to-video`, `gen4.5/image-to-video`. See the [Model Square](https://omniall.ai/pricing).

## Unified scenarios

Same endpoint `POST /v1/task/submit` — switch examples in the Create task page:

| Scenario | Key `input` fields |
| - | - |
| Text-to-video | `promptText` + optional `ratio` / `duration` |
| Image-to-video | `promptText` + `promptImage` |
| Video-to-video | `videoUri` + `promptText` (+ `ratio` / `duration`) |


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.