# Omniall AI

## Docs

### Getting started

- [Welcome to Omniall AI](https://docs.omniall.ai/introduction.md): Omniall AI API gateway docs: quickstart, model APIs, and Creation Center.
- [Quickstart](https://docs.omniall.ai/quickstart.md): Create an API key and call the OpenAI-compatible Chat Completions API.

### Guides

- [Guides overview](https://docs.omniall.ai/guides.md): Customer guides mapped to the Omniall AI API Reference.
- [Creation Center](https://docs.omniall.ai/creation-center.md): Use Omniall Creation Center for image, video, canvas, and Agent workflows.
- [Chat API](https://docs.omniall.ai/chat-api.md): Call ChatGPT, Claude, Gemini, and Responses through one Base URL.
- [Images API](https://docs.omniall.ai/images-api.md): Generate and edit images with OpenAI-style, Gemini native, and Midjourney paths.
- [File upload API](https://docs.omniall.ai/files-upload-api.md): Upload files via multipart and reuse the returned CDN URL.

#### Video

- [Video API](https://docs.omniall.ai/video-api.md): Async video generation overview across Omniall AI providers.
- [Google Omni video API](https://docs.omniall.ai/google-omni-video-api.md): Call Google Omni (omni-fast / omni-fast-v2v) via /v1/videos.
- [Veo video API](https://docs.omniall.ai/veo-video-api.md): Call Google Veo 3.1 through Omniall AI /v1/videos.
- [Grok video API](https://docs.omniall.ai/grok-video-api.md): Call xAI Grok Imagine Video via /v1/videos.
- [MiniMax-H3 video API](https://docs.omniall.ai/minimax-h3-video-api.md): MiniMax-H3 text/image/multimodal video, Context IR, 480P–2K, up to 30s, and 2K regeneration.
- [Doubao Seedance video API](https://docs.omniall.ai/doubao-seedance-video-api.md): Seedance 2.0 / 2.5 video task integration.
- [Kling video API](https://docs.omniall.ai/kling-video-api.md): Kling video generation through Omniall AI (OpenAI-style video tasks).
- [Runway video API](https://docs.omniall.ai/runway-video-api.md): Runway video via unified task API and official-style paths.
- [Wan video API](https://docs.omniall.ai/wan-video-api.md): Tongyi Wanxiang / Wan async video (wan2.1 480P/720P/1080P).

#### Audio

- [Audio API](https://docs.omniall.ai/audio-api.md): Music and speech APIs: Suno, MiniMax, and Google Lyria.
- [Audio speech (TTS) API](https://docs.omniall.ai/audio-speech-api.md): OpenAI-compatible speech API with MiniMax metadata extensions.
- [Suno music API](https://docs.omniall.ai/suno-music-api.md): Async Suno music and lyrics generation.
- [Lyria music API](https://docs.omniall.ai/lyria-music-api.md): Google Lyria music via Gemini generateContent.

## API Reference

### Overview

- [API overview](https://docs.omniall.ai/api-reference/overview.md): Omniall AI API reference overview: auth, endpoints, request/response shapes, and model selection.

### Models

- [List models](https://docs.omniall.ai/api-reference/endpoints/list-models.md): Returns models available to the authenticated API key / user group.

### Chat

#### ChatGPT

- [ChatGPT (Chat Completions)](https://docs.omniall.ai/api-reference/chatgpt.md): OpenAI Chat Completions format on Omniall AI.
- [Chat Completions](https://docs.omniall.ai/api-reference/endpoints/chat-completions.md): OpenAI Chat Completions API relayed by Omniall AI.

#### Claude

- [Claude API](https://docs.omniall.ai/api-reference/claude.md): Anthropic Messages API on Omniall AI.
- [Claude Messages](https://docs.omniall.ai/api-reference/endpoints/claude-messages.md): Anthropic Messages API relayed by Omniall AI.

#### Gemini

- [Gemini API](https://docs.omniall.ai/api-reference/gemini.md): Google Gemini generateContent APIs on Omniall AI.
- [Gemini generateContent](https://docs.omniall.ai/api-reference/endpoints/gemini-generate-content.md): Google Gemini `generateContent` relayed by Omniall AI.
- [Gemini streamGenerateContent](https://docs.omniall.ai/api-reference/endpoints/gemini-stream-generate-content.md): Google Gemini `streamGenerateContent` relayed by Omniall AI.

#### Responses

- [Responses](https://docs.omniall.ai/api-reference/responses.md): OpenAI Responses API format on Omniall AI.
- [Responses API](https://docs.omniall.ai/api-reference/endpoints/responses.md): OpenAI Responses API relayed by Omniall AI.
- [Responses compaction](https://docs.omniall.ai/api-reference/endpoints/responses-compact.md): Compact a long Responses conversation (OpenAI Responses compaction).

### Images

- [Images](https://docs.omniall.ai/api-reference/images.md): Painting / image generation APIs on Omniall AI.

#### General (recommended)

- [Generations](https://docs.omniall.ai/api-reference/endpoints/images-generations.md): OpenAI-compatible image generation.
- [Edit image](https://docs.omniall.ai/api-reference/endpoints/images-edits.md): OpenAI-compatible image edit / inpaint.

#### Gemini image

- [Overview](https://docs.omniall.ai/api-reference/image-gemini.md): Gemini image overview: Generations, native format, Chat image generation (text/image-to-image), Banana 2.1 PDF/video references, and Thinking levels.
- [Generations](https://docs.omniall.ai/api-reference/endpoints/images-generations.md): OpenAI-compatible image generation.
- [Gemini native format](https://docs.omniall.ai/api-reference/endpoints/gemini-native-image.md): Native Gemini `generateContent` for image generation.
- [Chat image generation](https://docs.omniall.ai/api-reference/endpoints/gemini-chat-image.md)

#### Jimeng / Seedream

- [Jimeng / Seedream image](https://docs.omniall.ai/api-reference/image-jimeng.md): Seedream (Jimeng) image generation via OpenAI Images.
- [Generations](https://docs.omniall.ai/api-reference/endpoints/images-generations.md): OpenAI-compatible image generation.

#### GPT-Image

- [GPT-Image](https://docs.omniall.ai/api-reference/image-gpt-image.md): OpenAI GPT-Image series via /v1/images.
- [Generations](https://docs.omniall.ai/api-reference/endpoints/images-generations.md): OpenAI-compatible image generation.
- [Edit image](https://docs.omniall.ai/api-reference/endpoints/images-edits.md): OpenAI-compatible image edit / inpaint.

#### Midjourney

- [Midjourney Imagine](https://docs.omniall.ai/api-reference/endpoints/mj-imagine.md): Create an Imagine task from a text prompt (and optional reference images in `base64Array`).
- [Midjourney Change](https://docs.omniall.ai/api-reference/endpoints/mj-change.md): Upscale / variation / reroll style change using `taskId` + `action`/`index` or `customId`.
- [Midjourney Simple Change](https://docs.omniall.ai/api-reference/endpoints/mj-simple-change.md): Simplified change API used by some MJ proxy clients.
- [Midjourney Describe](https://docs.omniall.ai/api-reference/endpoints/mj-describe.md): Describe an image as prompts (`base64Array`).
- [Midjourney Blend](https://docs.omniall.ai/api-reference/endpoints/mj-blend.md): Blend 2–5 images from `base64Array`.
- [Midjourney Edits](https://docs.omniall.ai/api-reference/endpoints/mj-edits.md): MJ edits / inpaint-style submit (`maskBase64` when needed).
- [Midjourney Action](https://docs.omniall.ai/api-reference/endpoints/mj-action.md): Execute a button action via `customId` from a previous task.
- [Midjourney Modal](https://docs.omniall.ai/api-reference/endpoints/mj-modal.md): Submit modal / remix follow-up content for a task.
- [Midjourney fetch task](https://docs.omniall.ai/api-reference/endpoints/mj-fetch.md): Fetch Midjourney task status/progress/result image URL.

#### Flux image

- [Flux image](https://docs.omniall.ai/api-reference/image-flux.md): Flux models via OpenAI Images generations.
- [Generations](https://docs.omniall.ai/api-reference/endpoints/images-generations.md): OpenAI-compatible image generation.

#### Grok Imagine

- [Grok Imagine](https://docs.omniall.ai/api-reference/image-grok.md): Grok image models via OpenAI Images.
- [Generations](https://docs.omniall.ai/api-reference/endpoints/images-generations.md): OpenAI-compatible image generation.

### Video

- [Video](https://docs.omniall.ai/api-reference/videos.md): Omniall AI video generation API overview.

#### Unified

- [Create video task](https://docs.omniall.ai/api-reference/endpoints/videos-create.md): Asynchronous video task. Poll GET /v1/videos/{task_id} until complete. Body fields vary by model family (see Docs guides).
- [Get video task](https://docs.omniall.ai/api-reference/endpoints/videos-get.md)
- [Create video generation](https://docs.omniall.ai/api-reference/endpoints/video-generations-create.md): Alternate async create path used by some model families (e.g. Seedance).
- [Get video generation](https://docs.omniall.ai/api-reference/endpoints/video-generations-get.md)

#### Google Omni

- [Google Omni](https://docs.omniall.ai/api-reference/video-google-omni.md): Google Omni async video (omni-fast / omni-fast-v2v).
- [Create video](https://docs.omniall.ai/api-reference/endpoints/omni-create.md): Create a Google Omni video task.
- [Get result](https://docs.omniall.ai/api-reference/endpoints/omni-get.md): Poll Omni video task.

#### Veo

- [Veo](https://docs.omniall.ai/api-reference/video-veo.md): Google Veo 3.1 async video API.
- [Create video](https://docs.omniall.ai/api-reference/endpoints/veo-create.md): Create a Veo 3.1 video task.
- [Get task status](https://docs.omniall.ai/api-reference/endpoints/veo-get.md): Poll Veo task.

#### Grok video

- [Grok video](https://docs.omniall.ai/api-reference/video-grok.md): xAI Grok Imagine video API.
- [Create video](https://docs.omniall.ai/api-reference/endpoints/grok-video-create.md): Grok Imagine async video create.
- [Get result](https://docs.omniall.ai/api-reference/endpoints/grok-video-get.md): Fetch Grok video task status/result.

#### MiniMax-H3

- [MiniMax-H3](https://docs.omniall.ai/api-reference/video-minimax-h3.md): MiniMax-H3 text/image/multimodal video, Context IR, 480P–2K, up to 30s, and 768P→2K regeneration.
- [Create video](https://docs.omniall.ai/api-reference/endpoints/minimax-h3-create.md): MiniMax-H3 async video create.
- [H3 Context IR create](https://docs.omniall.ai/api-reference/endpoints/minimax-h3-context-ir.md): MiniMax-H3 **Context IR** enhanced create.
- [Create regeneration task](https://docs.omniall.ai/api-reference/endpoints/minimax-h3-regeneration.md): Upscale a successful MiniMax-H3 **768P** video to **2K**.
- [Get result](https://docs.omniall.ai/api-reference/endpoints/minimax-h3-get.md): Fetch MiniMax-H3 task status/result.

#### Seedance

- [Seedance](https://docs.omniall.ai/api-reference/video-seedance.md): ByteDance Seedance 2.0 / 2.5 video API (combined).
- [Create video](https://docs.omniall.ai/api-reference/endpoints/seedance-create.md): Seedance (2.0 / 2.5) async video create.
- [Get result](https://docs.omniall.ai/api-reference/endpoints/seedance-get.md): Fetch Seedance video task status/result.

#### Kling

- [Kling](https://docs.omniall.ai/api-reference/video-kling.md): Kling video via OpenAI-style /v1/videos.
- [Create video](https://docs.omniall.ai/api-reference/endpoints/kling-create.md): Create a Kling video task.
- [Get task status](https://docs.omniall.ai/api-reference/endpoints/kling-get.md): Poll Kling task status.
- [Download video](https://docs.omniall.ai/api-reference/endpoints/kling-content.md): Download video binary after `status` is `completed`.

#### Runway

- [Runway](https://docs.omniall.ai/api-reference/video-runway.md): Runway video via unified task API and official-style paths.

##### Unified (recommended)

- [Create task](https://docs.omniall.ai/api-reference/endpoints/runway-unified-create.md): Submit a Runway task (text / image / video to video).
- [Get task result](https://docs.omniall.ai/api-reference/endpoints/runway-unified-get.md): Poll unified Runway task.

##### Official format

- [Image to video (official)](https://docs.omniall.ai/api-reference/endpoints/runway-official-i2v.md): Official-style image-to-video.
- [Video to video (official)](https://docs.omniall.ai/api-reference/endpoints/runway-official-v2v.md): Official-style video-to-video.
- [Get task (official)](https://docs.omniall.ai/api-reference/endpoints/runway-official-get.md): Poll official-style Runway task.

#### Wan video

- [Wan video](https://docs.omniall.ai/api-reference/video-wan.md): Tongyi Wanxiang / Wan async video API.
- [Create video](https://docs.omniall.ai/api-reference/endpoints/wan-create.md): Wan (Tongyi Wanxiang) async video create.
- [Get result](https://docs.omniall.ai/api-reference/endpoints/wan-get.md): Fetch Wan video task status/result.

### Audio

- [Audio](https://docs.omniall.ai/api-reference/audio.md): Omniall AI audio & music API overview.

#### Suno music

- [Suno music](https://docs.omniall.ai/api-reference/audio-suno.md): Suno async music and lyrics API.
- [Generate music](https://docs.omniall.ai/api-reference/endpoints/suno-music-create.md): Create a Suno music task.
- [Generate lyrics](https://docs.omniall.ai/api-reference/endpoints/suno-lyrics-create.md): Create a Suno lyrics task.
- [Get task [song]](https://docs.omniall.ai/api-reference/endpoints/suno-fetch-song.md): Get a single Suno task by public id.
- [Get task [lyrics]](https://docs.omniall.ai/api-reference/endpoints/suno-fetch-lyrics.md): Get a single Suno task by public id.
- [Batch get tasks](https://docs.omniall.ai/api-reference/endpoints/suno-fetch-batch.md): Batch fetch Suno tasks by public ids.

#### MiniMax audio

- [MiniMax audio](https://docs.omniall.ai/api-reference/audio-minimax.md): MiniMax music generation and speech synthesis.
- [Music generation](https://docs.omniall.ai/api-reference/endpoints/minimax-music-create.md): MiniMax official music generation (sync-style JSON response).
- [Speech synthesis](https://docs.omniall.ai/api-reference/endpoints/minimax-speech-create.md): Text-to-speech compatible with OpenAI Audio Speech.

#### Google Lyria

- [Google Lyria](https://docs.omniall.ai/api-reference/audio-lyria.md): Google Lyria music via Gemini generateContent.
- [Google native format](https://docs.omniall.ai/api-reference/endpoints/lyria-generate-content.md): Generate music with Google Lyria using Gemini `generateContent`.

### Platform

- [Platform](https://docs.omniall.ai/api-reference/platform.md): Platform utilities such as file upload.
- [Upload file](https://docs.omniall.ai/api-reference/endpoints/files-upload.md): Upload a single file via multipart form.

### Rerank

- [Rerank documents](https://docs.omniall.ai/api-reference/endpoints/rerank.md): Rank documents by relevance to a query (Jina-style / OpenAI-compatible rerank).

## OpenAPI Specs

- [xgapi-public](/docs/openapi/xgapi-public.yaml)
- [gemini-native-image](/docs/openapi/gemini-native-image.yaml)
- [google-omni-video](/docs/openapi/google-omni-video.yaml)
- [veo-video](/docs/openapi/veo-video.yaml)
- [grok-video](/docs/openapi/grok-video.yaml)
- [minimax-h3-video](/docs/openapi/minimax-h3-video.yaml)
- [seedance-video](/docs/openapi/seedance-video.yaml)
- [kling-video](/docs/openapi/kling-video.yaml)
- [runway-unified](/docs/openapi/runway-unified.yaml)
- [runway-official](/docs/openapi/runway-official.yaml)
- [wan-video](/docs/openapi/wan-video.yaml)
- [suno-audio](/docs/openapi/suno-audio.yaml)
- [minimax-music](/docs/openapi/minimax-music.yaml)
- [minimax-speech](/docs/openapi/minimax-speech.yaml)
- [lyria-audio](/docs/openapi/lyria-audio.yaml)
- [files-upload](/docs/openapi/files-upload.yaml)
- [gemini-image-to-image](/docs/openapi/gemini-image-to-image.yaml)
- [gemini-text-to-image](/docs/openapi/gemini-text-to-image.yaml)
- [gemini-image-to-image](/docs/zh-CN/openapi/gemini-image-to-image.yaml)
- [gemini-text-to-image](/docs/zh-CN/openapi/gemini-text-to-image.yaml)

> The links below point to documentation indexes. Follow each `/_llms/` index recursively until you reach documentation pages.

## Indexes

- [Simplified Chinese (101 pages)](https://docs.omniall.ai/_llms/zh-cn.md): Documentation for Simplified Chinese.


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