{
  "schema_version": 1,
  "page": 1,
  "page_size": 50,
  "total": 42,
  "total_pages": 1,
  "items": [
    {
      "id": "2026-09-18-gpt-image-tiers",
      "published_at": "2026-09-18",
      "type": "model_launch",
      "modality": [
        "image"
      ],
      "title": {
        "en": "GPT Image 2.5 Flare & Sunburst Tier Routes Live",
        "zh": "GPT Image 2.5 Flare 与 Sunburst 档位路由上线"
      },
      "summary": {
        "en": "Dedicated tier routes `gpt-image-2.5-flare` and `gpt-image-2.5-sunburst` are now available for volume image generation on OmniMux.",
        "zh": "图像生成特惠档模型 `gpt-image-2.5-flare` 与低价档模型 `gpt-image-2.5-sunburst` 现已接入 OmniMux 提供生图调用。"
      },
      "models": [
        "gpt-image-2.5-flare",
        "gpt-image-2.5-sunburst"
      ],
      "tags": [
        "image",
        "openai"
      ],
      "links": [
        {
          "label": {
            "en": "GPT Image Reference",
            "zh": "GPT Image 接口文档"
          },
          "href": {
            "en": "/en/api-reference/image-series/gpt-image/generate",
            "zh": "/zh/api-reference/image-series/gpt-image/generate"
          }
        }
      ],
      "body": {
        "en": "## GPT Image Tiers\n\nOmniMux now supports dedicated tier routing models `gpt-image-2.5-flare` and `gpt-image-2.5-sunburst` under the standard image generation endpoint.\n\n### Model specification\n\n- **Endpoint:** `POST /v1/images/generations`\n- **Authentication:** Bearer token (`sk-...`)\n- **Tier models:** `gpt-image-2.5-flare` (discounted tier), `gpt-image-2.5-sunburst` (economy tier)\n- **Parameters:** Fully compatible with baseline `gpt-image-2.5`\n\n### Request example\n\n```bash\ncurl --request POST \\\n  --url https://api.omnimux.ai/v1/images/generations \\\n  --header 'Authorization: Bearer sk-...' \\\n  --header 'Content-Type: application/json' \\\n  --data '{\n    \"model\": \"gpt-image-2.5-flare\",\n    \"prompt\": \"A clean studio product photograph on white background\",\n    \"n\": 1\n  }'\n```\n\nParameter details are documented in the [GPT Image Reference](/en/api-reference/image-series/gpt-image/generate).",
        "zh": "## GPT Image 计费档位模型\n\nOmniMux 标准生图端点现已支持专属档位模型 `gpt-image-2.5-flare` 与 `gpt-image-2.5-sunburst`。\n\n### 接口规范\n\n- **调用端点：** `POST /v1/images/generations`\n- **鉴权方式：** Bearer 令牌鉴权（`sk-...`）\n- **档位模型：** `gpt-image-2.5-flare`（特惠档）、`gpt-image-2.5-sunburst`（经济档）\n- **调用参数：** 与基准版 `gpt-image-2.5` 保持完全一致\n\n### 请求示例\n\n```bash\ncurl --request POST \\\n  --url https://api.omnimux.ai/v1/images/generations \\\n  --header 'Authorization: Bearer sk-...' \\\n  --header 'Content-Type: application/json' \\\n  --data '{\n    \"model\": \"gpt-image-2.5-flare\",\n    \"prompt\": \"白底商品棚拍静物摄影图\",\n    \"n\": 1\n  }'\n```\n\n完整参数规范请参阅 [GPT Image 接口文档](/zh/api-reference/image-series/gpt-image/generate)。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-18-video-task-models",
      "published_at": "2026-09-18",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "Video Per-Call Task Channels and H3 Variants Available",
        "zh": "视频按次生成任务通道与 H3 衍生变体模型上线"
      },
      "summary": {
        "en": "Per-call fixed billing task models `seedance-2-0-task`, `seedance-2-5-task`, `minimax-h3-task`, along with variants `minimax-h3-turbo` and `minimax-h3-video`, are now live on OmniMux.",
        "zh": "按次计费视频任务模型 `seedance-2-0-task`、`seedance-2-5-task`、`minimax-h3-task` 及 H3 系列变体 `minimax-h3-turbo`、`minimax-h3-video` 现已接入 OmniMux 提供异步调用。"
      },
      "models": [
        "seedance-2-0-task",
        "seedance-2-5-task",
        "minimax-h3-task",
        "minimax-h3-turbo",
        "minimax-h3-video"
      ],
      "tags": [
        "video",
        "task",
        "seedance",
        "minimax"
      ],
      "links": [
        {
          "label": {
            "en": "Seedance 2.0 Reference",
            "zh": "Seedance 2.0 接口文档"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/seedance-2-0",
            "zh": "/zh/api-reference/video-series/models/seedance-2-0"
          }
        },
        {
          "label": {
            "en": "MiniMax H3 Reference",
            "zh": "MiniMax H3 接口文档"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/minimax-h3",
            "zh": "/zh/api-reference/video-series/models/minimax-h3"
          }
        }
      ],
      "body": {
        "en": "## Video Task and Variant Models\n\nOmniMux now supports dedicated per-call video task channels and H3 model variants under the standard video endpoint.\n\n### Model specification\n\n- **Endpoint:** `POST /v1/video/generations`\n- **Polling endpoint:** `GET /v1/video/generations/{task_id}`\n- **Per-call models:** `seedance-2-0-task`, `seedance-2-5-task`, `minimax-h3-task` (billed per task call)\n- **H3 variants:** `minimax-h3-turbo` (turbo rendering), `minimax-h3-video` (dedicated video pipeline)\n\n### Request example\n\n```json\n{\n  \"model\": \"seedance-2-0-task\",\n  \"prompt\": \"A cinematic drone shot of coastal mountains at sunset\",\n  \"seconds\": 5\n}\n```\n\nFull parameter definitions are documented in the [Seedance 2.0 Reference](/en/api-reference/video-series/models/seedance-2-0) and [MiniMax H3 Reference](/en/api-reference/video-series/models/minimax-h3).",
        "zh": "## 视频按次任务通道与变体模型\n\nOmniMux 标准视频端点现已支持按次任务计费模型与 H3 系列衍生变体。\n\n### 接口规范\n\n- **创建端点：** `POST /v1/video/generations`\n- **轮询端点：** `GET /v1/video/generations/{task_id}`\n- **按次计费模型：** `seedance-2-0-task`、`seedance-2-5-task`、`minimax-h3-task`（固定单次扣费）\n- **H3 衍生变体：** `minimax-h3-turbo`（极速出片）、`minimax-h3-video`（高画质专线）\n\n### 请求示例\n\n```json\n{\n  \"model\": \"seedance-2-0-task\",\n  \"prompt\": \"海边山脉日落航拍全景镜头\",\n  \"seconds\": 5\n}\n```\n\n完整参数定义请参阅 [Seedance 2.0 接口文档](/zh/api-reference/video-series/models/seedance-2-0) 与 [MiniMax H3 接口文档](/zh/api-reference/video-series/models/minimax-h3)。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-18-volc-doubao-asr",
      "published_at": "2026-09-18",
      "type": "model_launch",
      "modality": [
        "audio"
      ],
      "title": {
        "en": "Doubao ASR Speech Recognition Series Available",
        "zh": "豆包语音识别大模型系列接口上线"
      },
      "summary": {
        "en": "Speech-to-text models `doubao-asr-bigmodel`, `seedasr-auc`, and `bigasr-auc` are now available on OmniMux via POST /v1/audio/transcriptions.",
        "zh": "语音识别模型 `doubao-asr-bigmodel`、`seedasr-auc` 与 `bigasr-auc` 现已接入 OmniMux，支持通过 POST /v1/audio/transcriptions 端点进行录音转写。"
      },
      "models": [
        "doubao-asr-bigmodel",
        "seedasr-auc",
        "bigasr-auc"
      ],
      "tags": [
        "audio",
        "asr",
        "doubao",
        "volcengine"
      ],
      "links": [
        {
          "label": {
            "en": "Doubao ASR Reference",
            "zh": "豆包语音识别文档"
          },
          "href": {
            "en": "/en/api-reference/audio-series/models/doubao-asr",
            "zh": "/zh/api-reference/audio-series/models/doubao-asr"
          }
        }
      ],
      "body": {
        "en": "## Doubao Speech Recognition\n\nThe `doubao-asr-bigmodel`, `seedasr-auc`, and `bigasr-auc` models are now supported on OmniMux.\n\n### Model specification\n\n- **Model IDs:** `doubao-asr-bigmodel`, `seedasr-auc`, `bigasr-auc`\n- **Protocol:** OpenAI-compatible Audio Transcriptions (`POST /v1/audio/transcriptions`)\n- **Authentication:** Bearer token (`sk-...`)\n- **Input format:** Audio file multipart upload (`wav`, `mp3`, `ogg`, `m4a`)\n\n### Request example\n\n```bash\ncurl --request POST \\\n  --url https://api.omnimux.ai/v1/audio/transcriptions \\\n  --header 'Authorization: Bearer sk-...' \\\n  --header 'Content-Type: multipart/form-data' \\\n  --form 'file=@audio.mp3' \\\n  --form 'model=doubao-asr-bigmodel'\n```\n\nFull parameter specifications are documented in the [Doubao ASR Reference](/en/api-reference/audio-series/models/doubao-asr).",
        "zh": "## 豆包语音识别\n\n`doubao-asr-bigmodel`、`seedasr-auc` 与 `bigasr-auc` 语音转文字模型现已在 OmniMux 正式提供调用。\n\n### 接口规范\n\n- **模型标识：** `doubao-asr-bigmodel`, `seedasr-auc`, `bigasr-auc`\n- **调用协议：** 兼容 OpenAI 语音转文字（`POST /v1/audio/transcriptions`）\n- **鉴权方式：** Bearer 令牌鉴权（`sk-...`）\n- **音频格式：** 支持通过分段上传提交 `wav`, `mp3`, `ogg`, `m4a` 等常见格式\n\n### 请求示例\n\n```bash\ncurl --request POST \\\n  --url https://api.omnimux.ai/v1/audio/transcriptions \\\n  --header 'Authorization: Bearer sk-...' \\\n  --header 'Content-Type: multipart/form-data' \\\n  --form 'file=@audio.mp3' \\\n  --form 'model=doubao-asr-bigmodel'\n```\n\n完整参数定义详见 [豆包语音识别参考文档](/zh/api-reference/audio-series/models/doubao-asr)。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-18-volc-seed-audio",
      "published_at": "2026-09-18",
      "type": "model_launch",
      "modality": [
        "audio"
      ],
      "title": {
        "en": "VolcEngine Seed Audio (seed-audio-1.0) Available via TTS Endpoint",
        "zh": "火山引擎 Seed Audio (seed-audio-1.0) 语音生成接口上线"
      },
      "summary": {
        "en": "VolcEngine / BytePlus audio creation model `seed-audio-1.0` is now available on OmniMux via POST /v1/audio/speech.",
        "zh": "火山引擎 / BytePlus 音频创作模型 `seed-audio-1.0` 现已接入 OmniMux，支持通过 POST /v1/audio/speech 端点进行语音与音频合成。"
      },
      "models": [
        "seed-audio-1.0"
      ],
      "tags": [
        "audio",
        "tts",
        "volcengine",
        "byteplus"
      ],
      "links": [
        {
          "label": {
            "en": "Seed Audio Reference",
            "zh": "Seed Audio 接口文档"
          },
          "href": {
            "en": "/en/api-reference/audio-series/models/seed-audio",
            "zh": "/zh/api-reference/audio-series/models/seed-audio"
          }
        }
      ],
      "body": {
        "en": "## Seed Audio 1.0\n\nThe `seed-audio-1.0` model is now supported on OmniMux under standard audio speech endpoints.\n\n### Model specification\n\n- **Model ID:** `seed-audio-1.0`\n- **Protocol:** OpenAI-compatible Audio Speech (`POST /v1/audio/speech`)\n- **Authentication:** Bearer token (`sk-...`)\n- **Output limit:** Up to 120 seconds per generation\n- **Prompt limit:** Up to 3000 characters\n\n### Request example\n\n```bash\ncurl --request POST \\\n  --url https://api.omnimux.ai/v1/audio/speech \\\n  --header 'Authorization: Bearer sk-...' \\\n  --header 'Content-Type: application/json' \\\n  --data '{\n    \"model\": \"seed-audio-1.0\",\n    \"input\": \"Text to synthesize into speech\",\n    \"voice\": \"zh_female_cancan\",\n    \"response_format\": \"mp3\"\n  }' \\\n  --output output.mp3\n```\n\nFull parameter specifications are documented in the [Seed Audio Reference](/en/api-reference/audio-series/models/seed-audio).",
        "zh": "## Seed Audio 1.0\n\n火山引擎 `seed-audio-1.0` 语音生成大模型已接入 OmniMux 标准语音合成端点。\n\n### 接口规范\n\n- **模型标识：** `seed-audio-1.0`\n- **调用协议：** 兼容 OpenAI 语音合成（`POST /v1/audio/speech`）\n- **鉴权方式：** Bearer 令牌鉴权（`sk-...`）\n- **生成上限：** 单次合成最长 120 秒\n- **文本上限：** 提示文本最长 3000 字符\n\n### 请求示例\n\n```bash\ncurl --request POST \\\n  --url https://api.omnimux.ai/v1/audio/speech \\\n  --header 'Authorization: Bearer sk-...' \\\n  --header 'Content-Type: application/json' \\\n  --data '{\n    \"model\": \"seed-audio-1.0\",\n    \"input\": \"要合成的文本内容\",\n    \"voice\": \"zh_female_cancan\",\n    \"response_format\": \"mp3\"\n  }' \\\n  --output output.mp3\n```\n\n完整参数定义与支持格式请参阅 [Seed Audio 接口参考文档](/zh/api-reference/audio-series/models/seed-audio)。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-18-indextts-2",
      "published_at": "2026-09-18",
      "type": "model_launch",
      "modality": [
        "audio"
      ],
      "title": {
        "en": "Index TTS Voice Clone (indextts-2) Available",
        "zh": "Index TTS 声音克隆 (indextts-2) 接口上线"
      },
      "summary": {
        "en": "Dedicated voice clone and emotional TTS model `indextts-2` is now available on OmniMux via POST /v1/tasks/autodl.",
        "zh": "声音克隆与情绪语音合成模型 `indextts-2` 现已接入 OmniMux，支持通过 POST /v1/tasks/autodl 端点进行异步调用。"
      },
      "models": [
        "indextts-2"
      ],
      "tags": [
        "audio",
        "voice-clone",
        "tts"
      ],
      "links": [
        {
          "label": {
            "en": "Index TTS Reference",
            "zh": "Index TTS 接口参考"
          },
          "href": {
            "en": "/en/api-reference/audio-series/models/index-tts",
            "zh": "/zh/api-reference/audio-series/models/index-tts"
          }
        }
      ],
      "body": {
        "en": "## Index TTS 2\n\nThe `indextts-2` voice clone model is now supported on OmniMux.\n\n### Model specification\n\n- **Model ID:** `indextts-2`\n- **Create endpoint:** `POST /v1/tasks/autodl`\n- **Polling endpoint:** `GET /v1/tasks/{task_id}`\n- **Artifact endpoint:** `GET /v1/tasks/{task_id}/artifacts`\n- **Authentication:** Bearer token (`sk-...`)\n\n### Request example\n\n```json\n{\n  \"model\": \"indextts-2\",\n  \"prompt_text\": \"Hello, this is a cloned voice generation test.\",\n  \"prompt_simple\": \"https://example.com/reference_audio.wav\",\n  \"emo_control_method\": \"same_as_reference\"\n}\n```\n\nFull parameter specifications are documented in the [Index TTS Reference](/en/api-reference/audio-series/models/index-tts).",
        "zh": "## Index TTS 2\n\n声音克隆模型 `indextts-2` 现已接入 OmniMux 提供异步调用。\n\n### 接口规范\n\n- **模型标识：** `indextts-2`\n- **任务提交端点：** `POST /v1/tasks/autodl`\n- **任务状态查询：** `GET /v1/tasks/{task_id}`\n- **生成产物获取：** `GET /v1/tasks/{task_id}/artifacts`\n- **鉴权方式：** Bearer 令牌鉴权（`sk-...`）\n\n### 请求示例\n\n```json\n{\n  \"model\": \"indextts-2\",\n  \"prompt_text\": \"你好，这是一段声音克隆合成测试台词。\",\n  \"prompt_simple\": \"https://example.com/reference_audio.wav\",\n  \"emo_control_method\": \"与音色参考音频相同\"\n}\n```\n\n完整参数定义与情绪调节字段详见 [Index TTS 接口参考文档](/zh/api-reference/audio-series/models/index-tts)。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-18-suno-sounds",
      "published_at": "2026-09-18",
      "type": "model_launch",
      "modality": [
        "audio"
      ],
      "title": {
        "en": "Suno Sound Effects Generation (suno-sounds) Available",
        "zh": "Suno 独立音效生成接口 (suno-sounds) 上线"
      },
      "summary": {
        "en": "Dedicated sound effects generation model `suno-sounds` is now available on OmniMux via POST /v1/video/generations.",
        "zh": "独立音效生成模型 `suno-sounds` 现已接入 OmniMux，支持通过 POST /v1/video/generations 端点生成场景音效。"
      },
      "models": [
        "suno-sounds"
      ],
      "tags": [
        "audio",
        "sound-effects",
        "suno"
      ],
      "links": [
        {
          "label": {
            "en": "Suno API Reference",
            "zh": "Suno 接口文档"
          },
          "href": {
            "en": "/en/api-reference/audio-series/models/suno",
            "zh": "/zh/api-reference/audio-series/models/suno"
          }
        }
      ],
      "body": {
        "en": "## Suno Sound Effects\n\nThe dedicated sound effect generation model `suno-sounds` is now supported on OmniMux.\n\n### Model specification\n\n- **Model ID:** `suno-sounds`\n- **Endpoint:** `POST /v1/video/generations`\n- **Polling endpoint:** `GET /v1/video/generations/{task_id}`\n- **Authentication:** Bearer token (`sk-...`)\n- **Modality:** Ambient and scene sound effects without vocal lyrics\n\n### Request example\n\n```bash\ncurl --request POST \\\n  --url https://api.omnimux.ai/v1/video/generations \\\n  --header 'Authorization: Bearer sk-...' \\\n  --header 'Content-Type: application/json' \\\n  --data '{\n    \"model\": \"suno-sounds\",\n    \"prompt\": \"footsteps on gravel followed by heavy rain and distant thunder\"\n  }'\n```\n\nDetails are available in the [Suno API Reference](/en/api-reference/audio-series/models/suno).",
        "zh": "## Suno 音效生成\n\n独立场景音效生成模型 `suno-sounds` 现已在 OmniMux 提供异步调用支持。\n\n### 接口规范\n\n- **模型标识：** `suno-sounds`\n- **创建端点：** `POST /v1/video/generations`\n- **任务轮询端点：** `GET /v1/video/generations/{task_id}`\n- **鉴权方式：** Bearer 令牌鉴权（`sk-...`）\n- **生成模态：** 纯场景音效与环境音（无声乐伴奏与人声歌词）\n\n### 请求示例\n\n```bash\ncurl --request POST \\\n  --url https://api.omnimux.ai/v1/video/generations \\\n  --header 'Authorization: Bearer sk-...' \\\n  --header 'Content-Type: application/json' \\\n  --data '{\n    \"model\": \"suno-sounds\",\n    \"prompt\": \"碎石路上的脚步声，伴随暴雨和远处的雷声\"\n  }'\n```\n\n完整参数定义与状态说明请参阅 [Suno 接口参考文档](/zh/api-reference/audio-series/models/suno)。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-18-gemini-3-8-flash",
      "published_at": "2026-09-18",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "Gemini 3.8 Flash Available via Chat Completions",
        "zh": "Gemini 3.8 Flash 对话补全接口上线"
      },
      "summary": {
        "en": "Gemini 3.8 Flash (model `gemini-3.8-flash`) is now available on OmniMux via OpenAI-compatible Chat Completions.",
        "zh": "Gemini 3.8 Flash（模型标识 `gemini-3.8-flash`）现已通过兼容 OpenAI 的对话补全接口在 OmniMux 正式提供调用。"
      },
      "models": [
        "gemini-3.8-flash"
      ],
      "tags": [
        "gemini",
        "google",
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "Gemini Complete Reference",
            "zh": "Gemini 完整参数参考"
          },
          "href": {
            "en": "/en/api-reference/text-series/gemini/complete",
            "zh": "/zh/api-reference/text-series/gemini/complete"
          }
        }
      ],
      "body": {
        "en": "## Gemini 3.8 Flash\n\nThe `gemini-3.8-flash` model is now supported on OmniMux under the standard Gemini Chat Completions endpoint.\n\n### Model specification\n\n- **Model ID:** `gemini-3.8-flash`\n- **Protocol:** OpenAI-compatible Chat Completions\n- **Endpoint:** `POST /v1/chat/completions`\n- **Authentication:** Bearer token (`sk-...`)\n- **Streaming:** Supported via `\"stream\": true`\n\n### Request example\n\n```json\n{\n  \"model\": \"gemini-3.8-flash\",\n  \"messages\": [\n    {\n      \"role\": \"user\",\n      \"content\": \"Hello\"\n    }\n  ]\n}\n```\n\nFull parameter schema and status codes are documented in the [Gemini Complete API Reference](/en/api-reference/text-series/gemini/complete).",
        "zh": "## Gemini 3.8 Flash\n\n`gemini-3.8-flash` 模型已接入 OmniMux 标准对话补全服务。\n\n### 接口规范\n\n- **模型标识：** `gemini-3.8-flash`\n- **接口协议：** 兼容 OpenAI 对话补全协议\n- **调用端点：** `POST /v1/chat/completions`\n- **鉴权方式：** Bearer 令牌鉴权（`sk-...`）\n- **流式传输：** 支持通过 `\"stream\": true` 启用\n\n### 请求体示例\n\n```json\n{\n  \"model\": \"gemini-3.8-flash\",\n  \"messages\": [\n    {\n      \"role\": \"user\",\n      \"content\": \"你好\"\n    }\n  ]\n}\n```\n\n完整参数定义与状态码说明请查阅 [Gemini 完整参数参考文档](/zh/api-reference/text-series/gemini/complete)。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-18-doubao-seed-evolving",
      "published_at": "2026-09-18",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "ByteDance Doubao official direct model seed-evolving is now live",
        "zh": "字节跳动火山方舟官方直连大模型 seed-evolving 正式上线"
      },
      "summary": {
        "en": "OmniMux adds official direct channel #73 for ByteDance Doubao seed-evolving. Features 1024K context window, 256K max output, deep reasoning, and transparent 1:1 parity pricing with billing expressions, fully verified via local OmniMux CLI.",
        "zh": "OmniMux 新增字节跳动火山方舟官方直连渠道 #73，正式上线豆包多模态推理大模型 seed-evolving。具备 1024K 上下文窗口、256K 最大输出、深度思考与多模态理解能力，按官方原价 1:1 表达式透明计量，已通过本地 OmniMux CLI 端到端验证可用。"
      },
      "models": [
        "seed-evolving"
      ],
      "tags": [
        "doubao",
        "seed-evolving",
        "llm",
        "bytedance",
        "volcengine",
        "cli"
      ],
      "links": [
        {
          "label": {
            "en": "Doubao Complete API Reference",
            "zh": "Doubao 完整参数手册"
          },
          "href": {
            "en": "/en/api-reference/text-series/doubao/complete",
            "zh": "/zh/api-reference/text-series/doubao/complete"
          }
        }
      ],
      "body": {
        "en": "## What is new\n\n- **Official Direct Channel**: Added VolcEngine Ark direct route `#73` for ByteDance Doubao `seed-evolving`.\n- **Full Multimodal Reasoning**: 1024K context window, 256K max output tokens, native deep thinking, tool calling, and structured outputs.\n- **Transparent 1:1 Parity Pricing**: Token billing expressions anchor official list prices ($0.8824/M input, $4.4118/M output, $0.1765/M cache hit; FX 6.8).\n- **Local OmniMux CLI Verification**: Verified live invocation and pricing queries using the local `omnimux` CLI (`omnimux models`, `omnimux pricing`, `omnimux tokens exec`).\n\n## Documentation\n\n- Doubao API Reference: `/en/api-reference/text-series/doubao/complete`",
        "zh": "## 本次更新\n\n- **官方直连专线**：新增火山方舟官方直连渠道 `#73`，直通字节跳动豆包 `seed-evolving` 模型。\n- **全模态深度思考**：提供 1024K 上下文窗口、256K 最大回答输出、原生深度思考、工具调用与多模态理解能力。\n- **透明官方同价**：计费表达式严格锚定官方刊例原价（输入 6.00 元/M、输出 30.00 元/M、缓存命中 1.20 元/M，按 6.8 汇率透明折算）。\n- **本地 OmniMux CLI 验证通过**：使用官方命令行工具 `omnimux` 完成模型可见性、定价核验与安全令牌端到端对话全流程实测验证。\n\n## 参考文档\n\n- Doubao 完整参数手册：`/zh/api-reference/text-series/doubao/complete`"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-14-image-reference-fields",
      "published_at": "2026-09-14",
      "type": "capability",
      "modality": [
        "image"
      ],
      "title": {
        "en": "Image model pages now document the reference-image field and its limits",
        "zh": "生图模型页补充参考图字段与上限说明"
      },
      "summary": {
        "en": "Image-generation model pages now list the reference-image request field `images` (with its accepted aliases) and state the limits each vendor actually publishes. A previously stated per-file size cap is withdrawn: the 50 MB figure is scoped to mask editing, not to reference images.",
        "zh": "生图模型页新增参考图请求字段 `images`（含可接受别名），并按模型给出厂商确实公布的上限。此前写出的单文件大小上限撤回：50 MB 属于遮罩编辑口径，不是参考图口径。"
      },
      "models": [
        "gpt-image-2.5",
        "mj-v7",
        "mj-v8-1",
        "nano-banana-2",
        "nano-banana-pro",
        "qwen-image-3-0",
        "seedream-5-0-pro"
      ],
      "tags": [
        "image",
        "reference-images",
        "docs"
      ],
      "links": [
        {
          "label": {
            "en": "Image series model pages",
            "zh": "图像系列模型页"
          },
          "href": {
            "en": "/en/api-reference/image-series/models/nano-banana-2",
            "zh": "/zh/api-reference/image-series/models/nano-banana-2"
          }
        },
        {
          "label": {
            "en": "GPT Image generation",
            "zh": "GPT Image 图像生成"
          },
          "href": {
            "en": "/en/api-reference/image-series/gpt-image/generate",
            "zh": "/zh/api-reference/image-series/gpt-image/generate"
          }
        },
        {
          "label": {
            "en": "Midjourney image generation",
            "zh": "Midjourney 生图"
          },
          "href": {
            "en": "/en/api-reference/image-series/models/midjourney",
            "zh": "/zh/api-reference/image-series/models/midjourney"
          }
        }
      ],
      "body": {
        "en": "## What changed\n\n- Image-generation model pages now document the reference-image request field `images` (`string[]`, image URL or base64) together with the accepted aliases `image`, `image_urls` and `input_reference`, which the gateway normalizes to `images` before the upstream call.\n- Each page states the limits its vendor publishes for that model: maximum count, minimum, accepted formats and per-file size ceiling.\n\n## Limits we do not state\n\n- Where a value is not published by the vendor, the page says so instead of carrying a number; nothing is rounded or copied from a sibling model.\n- Where no model spec exists yet, the reference-image note marks the capability **unverified** and states no limit at all.\n\n## Withdrawn figure\n\n- The previously stated per-file reference-image cap of 50 MB is withdrawn. In the vendor's image guide that figure sits under mask requirements and covers the image being edited together with its mask; it is not a reference-image ceiling. `gpt-image-2.5` therefore carries no per-file size limit.\n\n## Documentation\n\n- Image series model pages: `/en/api-reference/image-series/models/*`\n- GPT Image generation: `/en/api-reference/image-series/gpt-image/generate`",
        "zh": "## 本次变更\n\n- 生图模型页新增参考图请求字段 `images`（`string[]`，图片 URL 或 base64），并写明可接受的别名 `image`、`image_urls`、`input_reference`；网关会在调用上游前统一归一为 `images`。\n- 每页按厂商口径给出该模型的参考图上限：单次张数、下限、接受格式与单文件大小上限。\n\n## 不声明的上限\n\n- 厂商未公布的项直接写“官方未公布”，不给数字，也不从同系列其他模型套用或取整。\n- 尚无模型档案的模型，参考图说明标注为**未核验**，且完全不声明上限。\n\n## 撤回的数字\n\n- 此前写出的单张参考图 50 MB 上限撤回。厂商图像指南中的该数字位于遮罩（mask）要求下，覆盖的是“被编辑图 + 其遮罩”，不是参考图上限。因此 `gpt-image-2.5` 不声明单文件大小上限。\n\n## 文档\n\n- 图像系列模型页：`/zh/api-reference/image-series/models/*`\n- GPT Image 图像生成：`/zh/api-reference/image-series/gpt-image/generate`"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-13-midjourney-two-ids",
      "published_at": "2026-09-13",
      "type": "breaking",
      "modality": [
        "image"
      ],
      "title": {
        "en": "Midjourney now uses two version IDs with per-request actions",
        "zh": "Midjourney 收敛为两个版本 ID，动作改为按请求选择"
      },
      "summary": {
        "en": "Call model mj-v7 or mj-v8-1 and select the action with model_params.operation. The old per-action model IDs and the legacy `midjourney` ID are no longer accepted. Per-call prices are now tiered by action, and quality/speed/upscale-type multipliers apply on top.",
        "zh": "请改用模型 mj-v7 或 mj-v8-1，并通过 model_params.operation 选择动作。原有的按动作模型 ID 与旧的 `midjourney` ID 已不再接受。按次价格改为按动作分档，并在其上叠加画质/速度/放大类型乘子。"
      },
      "models": [
        "mj-v7",
        "mj-v8-1"
      ],
      "tags": [
        "image",
        "breaking",
        "migration",
        "pricing"
      ],
      "links": [
        {
          "label": {
            "en": "Midjourney image generation",
            "zh": "Midjourney 图像生成"
          },
          "href": {
            "en": "/en/api-reference/image-series/models/midjourney",
            "zh": "/zh/api-reference/image-series/models/midjourney"
          }
        }
      ],
      "body": {
        "en": "## Required migration\n\n- Replace the legacy `midjourney` ID and every per-action ID (for example `mj-v7-upscale`, `mj-v8-1-remove-bg`) with one of two version IDs: `mj-v7` or `mj-v8-1`.\n- Select the action in the request body: `model_params.operation` = `generate` | `enhance` | `pan` | `outpaint` | `remove-bg` | `upscale` | `variation` | `edit` | `inpaint` | `remix` | `retexture` | `upload-paint`. Omit it for plain generation.\n- Keep using `POST /v1/images/generations`; the request shape, the async task id and the polling behaviour are unchanged.\n\nThe old IDs are not compatibility aliases. An unknown or unpriced `operation` is rejected with **400**, never silently downgraded to plain generation.\n\n## Pricing\n\nPer-call prices are tiered by action, and the tiers are the same on both versions:\n\n| Action tier | Price | Credits |\n| --- | --- | --- |\n| `generate`, `enhance`, `pan`, `outpaint`, `remove-bg` | USD 0.113446 | 1.134 |\n| `upscale`, `variation`, `edit`, `inpaint`, `remix`, `retexture`, `upload-paint` | USD 0.170169 | 1.702 |\n\nMultipliers stack on the action tier: `quality: \"hd\"` (mj-v8-1) ×1.5, `model_params.speed: \"turbo\"` ×2, upscale `type: \"creative\"` ×4/3. `draft`/`fast` and `type: \"standard\"` stay ×1. Measured example: mj-v8-1 + hd + upscale + creative = USD 0.255251 per call.\n\nThe `mj-v7` per-call price moved from USD 0.056723 to USD 0.113446 to match what upstream bills for generation; the action tiers above are unchanged in absolute terms.\n\n## Inpaint\n\nUse the mask-image form: `model_params.mask.url` (an image URL or a `data:` URL, black = keep, white = regenerate). The polygon form `model_params.mask.areas[]` is rejected upstream with `invalid_parameters: Image dimensions out of range` in our measurements (2026-09-13), including with the upstream documentation's own example coordinates.",
        "zh": "## 必须迁移\n\n- 将旧的 `midjourney` ID 以及所有按动作的 ID（例如 `mj-v7-upscale`、`mj-v8-1-remove-bg`）替换为两个版本 ID 之一：`mj-v7` 或 `mj-v8-1`。\n- 动作改在请求体中选择：`model_params.operation` 取 `generate` | `enhance` | `pan` | `outpaint` | `remove-bg` | `upscale` | `variation` | `edit` | `inpaint` | `remix` | `retexture` | `upload-paint`；不传即为普通生成。\n- 继续使用 `POST /v1/images/generations`；请求形态、异步任务 id 与轮询方式均不变。\n\n旧 ID 不是兼容别名。未知或无法定价的 `operation` 一律返回 **400**，绝不静默降级为普通生成。\n\n## 价格\n\n按次价格改为按动作分档，两个版本的档位一致：\n\n| 动作档位 | 价格 | 积分 |\n| --- | --- | --- |\n| `generate`、`enhance`、`pan`、`outpaint`、`remove-bg` | USD 0.113446 | 1.134 |\n| `upscale`、`variation`、`edit`、`inpaint`、`remix`、`retexture`、`upload-paint` | USD 0.170169 | 1.702 |\n\n在动作档位之上叠加乘子：`quality: \"hd\"`（仅 mj-v8-1）×1.5、`model_params.speed: \"turbo\"` ×2、upscale `type: \"creative\"` ×4/3；`draft`/`fast` 与 `type: \"standard\"` 为 ×1。实测例：mj-v8-1 + hd + upscale + creative = USD 0.255251/次。\n\n`mj-v7` 的按次价由 USD 0.056723 调整为 USD 0.113446，与上游生成档实际扣费对齐；上述各动作档的绝对价不变。\n\n## 局部重绘\n\n请使用图片 mask：`model_params.mask.url`（图片 URL 或 `data:` URL，黑色=保留，白色=重绘）。多边形写法 `model_params.mask.areas[]` 在我们 2026-09-13 的实测中被上游以 `invalid_parameters: Image dimensions out of range` 拒绝，连上游文档自带的示例坐标也一样。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-12-console-playground-generation",
      "published_at": "2026-09-12",
      "type": "platform",
      "modality": [
        "platform"
      ],
      "title": {
        "en": "Model page playground generates through the gateway session",
        "zh": "模型详情页 Playground 支持以网关登录态直接发起生成"
      },
      "summary": {
        "en": "The public model page submits real generations from a signed-in browser session, so no separately minted API token is needed to try a model from the console.",
        "zh": "公开模型详情页现可直接以浏览器登录态提交真实生成，控制台试用模型无需另行创建 API 令牌。"
      },
      "models": [],
      "tags": [
        "console",
        "playground",
        "session",
        "video",
        "platform"
      ],
      "links": [
        {
          "label": {
            "en": "Video generation API",
            "zh": "视频生成 API"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/seedance-2-5",
            "zh": "/zh/api-reference/video-series/models/seedance-2-5"
          }
        }
      ],
      "body": {
        "en": "## Overview\n\nThe model page at `omnimux.ai/pricing/<model-id>` now runs real generations. Its **Playground** section submits through the gateway using the signed-in browser session, so a user who has already logged in to OmniMux can try a model without creating an API token first.\n\n## Which path to call\n\n- **From the console**: the page posts to the session-authenticated `POST /pg/video/generations`. This path accepts a browser session only and is not part of the token API surface.\n- **From your own code**: keep using `POST /v1/video/generations` with an API token — the same contract documented in the video-series model pages. Nothing about the token path changed.\n\n## Task results\n\nSubmitted tasks are read back through the same session: `GET /api/task/self` for status, progress and quota, and `GET /api/task/:task_id/artifacts` for the media projection. Tasks submitted from the console appear in the account's task list like any other task.\n\n## Billing\n\nConsole submissions use the standard precharge and settle path. No separate ledger, pricing rule or quota bucket is introduced for the console path.",
        "zh": "## 概述\n\n位于 `omnimux.ai/pricing/<模型 ID>` 的模型详情页现已支持真实生成。页面中的 **Playground** 区块以浏览器登录态经网关提交，因此已登录 OmniMux 的用户无需先创建 API 令牌即可试用模型。\n\n## 该调用哪条路径\n\n- **在控制台内**：页面调用会话鉴权的 `POST /pg/video/generations`。该路径只接受浏览器登录态，不属于令牌 API 面。\n- **在自己的代码里**：继续使用带 API 令牌的 `POST /v1/video/generations`，契约与视频系列模型页所述一致。令牌路径没有任何改动。\n\n## 任务结果\n\n提交后的任务同样以登录态读取：`GET /api/task/self` 取状态、进度与额度，`GET /api/task/:task_id/artifacts` 取媒体投影。控制台提交的任务会和其他任务一样出现在账户任务列表中。\n\n## 计费\n\n控制台提交走标准预扣与结算路径，不为控制台单独引入账本、定价规则或配额池。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-10-r2-file-storage",
      "published_at": "2026-09-10",
      "type": "platform",
      "modality": [
        "platform"
      ],
      "title": {
        "en": "Cloudflare R2 storage & Evolink-compatible file upload API released",
        "zh": "Cloudflare R2 存储开启与对齐 Evolink 文件上传 API 上线"
      },
      "summary": {
        "en": "Supports Base64, stream multipart, and remote URL uploads with 72h lifecycle and 2,000 files/user quota.",
        "zh": "支持 Base64、流式表单及 URL 远程抓取上传，配备 72 小时自动过期与单用户 2000 个文件配额门禁。"
      },
      "models": [],
      "tags": [
        "storage",
        "files",
        "r2",
        "api"
      ],
      "links": [
        {
          "label": {
            "en": "Base64 upload",
            "zh": "Base64 文件上传"
          },
          "href": {
            "en": "/en/api-reference/file-series/upload-base64",
            "zh": "/zh/api-reference/file-series/upload-base64"
          }
        },
        {
          "label": {
            "en": "File quota",
            "zh": "查询文件配额"
          },
          "href": {
            "en": "/en/api-reference/file-series/quota",
            "zh": "/zh/api-reference/file-series/quota"
          }
        }
      ],
      "body": {
        "en": "## Overview\n\nOmniMux now provides Cloudflare R2 object storage integration alongside a local disk fallback driver. File series endpoints are available under `/api/v1/files/*` and `/v1/files/*`.\n\n## Capabilities\n\n- **Base64 upload**: `POST /api/v1/files/upload/base64` supporting Data URL and raw Base64.\n- **Stream upload**: `POST /api/v1/files/upload/stream` via `multipart/form-data`.\n- **URL upload**: `POST /api/v1/files/upload/url` with SSRF protection.\n- **Quota & Expiration**: 2,000 active files per user quota, 72-hour automated expiration with background cleanup.",
        "zh": "## 概述\n\nOmniMux 现已全量支持 Cloudflare R2 对象存储与本地安全降级驱动，文件系列端点挂载于 `/api/v1/files/*` 与 `/v1/files/*`。\n\n## 核心能力\n\n- **Base64 上传**：`POST /api/v1/files/upload/base64` 支持 Data URL 与纯 Base64。\n- **流式上传**：`POST /api/v1/files/upload/stream` 支持 `multipart/form-data` 表单流。\n- **URL 远程抓取**：`POST /api/v1/files/upload/url` 内置 SSRF 防护。\n- **配额与 72h 过期**：单用户 2000 个活跃文件上限，72 小时自动过期物理清理。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-10-gpt-image-25-rename",
      "published_at": "2026-09-10",
      "type": "breaking",
      "modality": [
        "image"
      ],
      "title": {
        "en": "GPT Image per-call model IDs renamed to 2.5",
        "zh": "GPT Image 按次模型 ID 正式更名为 2.5"
      },
      "summary": {
        "en": "Update requests to gpt-image-2.5 or gpt-image-2.5-hd. Old model IDs are no longer accepted; per-call prices are unchanged.",
        "zh": "请将请求改为 gpt-image-2.5 或 gpt-image-2.5-hd。旧 model ID 不再接受，按次售价不变。"
      },
      "models": [
        "gpt-image-2.5",
        "gpt-image-2.5-hd"
      ],
      "tags": [
        "image",
        "breaking",
        "migration"
      ],
      "links": [
        {
          "label": {
            "en": "GPT Image generation",
            "zh": "GPT Image 图像生成"
          },
          "href": {
            "en": "/en/api-reference/image-series/gpt-image/generate",
            "zh": "/zh/api-reference/image-series/gpt-image/generate"
          }
        }
      ],
      "body": {
        "en": "## Required migration\n\n- Replace `gpt-image-2` with `gpt-image-2.5`.\n- Replace `gpt-image-2-hd` with `gpt-image-2.5-hd`.\n- Keep using `POST /v1/images/generations`. Update the `model` value in saved requests, integrations, and any model allowlists.\n\nThe old IDs are not compatibility aliases. The `gpt-image-2-5` spelling is not supported.\n\n## Pricing and scope\n\nBase per-call prices remain USD 0.0441 for `gpt-image-2.5` and USD 0.005479 for `gpt-image-2.5-hd`; group multipliers may affect final charges. This is a public ID migration, not a claim of a newly verified resolution or upstream capability.\n\nFlare and Sunburst are separate token-billed models and are not available as part of this change.",
        "zh": "## 必须迁移\n\n- 将 `gpt-image-2` 替换为 `gpt-image-2.5`。\n- 将 `gpt-image-2-hd` 替换为 `gpt-image-2.5-hd`。\n- 继续使用 `POST /v1/images/generations`，更新保存的请求、集成代码及模型白名单中的 `model`。\n\n旧 ID 不是兼容别名，也不支持 `gpt-image-2-5` 拼写。\n\n## 价格与范围\n\n基础按次价保持不变：`gpt-image-2.5` 为 USD 0.0441/次，`gpt-image-2.5-hd` 为 USD 0.005479/次；分组倍率可能影响最终费用。本次是公开 ID 迁移，不代表新增或验证了特定分辨率及上游能力。\n\nFlare 和 Sunburst 是独立的 Token 计费模型，不在本次变更中开放。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-09-h3-media-task-contracts",
      "published_at": "2026-09-09",
      "type": "breaking",
      "modality": [
        "video"
      ],
      "title": {
        "en": "H3 ordered references and task result handling",
        "zh": "H3 有序参考素材与任务结果处理修复"
      },
      "summary": {
        "en": "Explicit operations, ordered references and 720p migration; task reads preserve settlement ownership.",
        "zh": "明确操作类型、保留参考素材顺序并迁移720p；任务查询保留后台结算职责。"
      },
      "models": [
        "minimax/h3-max",
        "minimax/h3-max-turbo"
      ],
      "tags": [],
      "links": [
        {
          "label": {
            "en": "H3 Max API",
            "zh": "H3 Max API"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/minimax-h3-max",
            "zh": "/zh/api-reference/video-series/models/minimax-h3-max"
          }
        }
      ],
      "body": {
        "en": "Use operation to select text, first-frame or Max multi-reference generation. Media arrays retain order and conflicting aliases are rejected. Replace legacy 720p with explicit 768P. Temporary result unavailability is retried; failures and cancellations are not masked by stale video URLs. Status reads no longer persist terminal transitions ahead of settlement polling.",
        "zh": "使用 operation 选择文生视频、首帧或 Max 多参考生成。素材数组保留顺序，冲突同义字段会被拒绝。旧720p需明确改为768P。产物暂不可用时重试，失败或取消不再被旧视频URL掩盖。状态查询不再抢先持久化终态，结算仍由后台轮询负责。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-08-fal-minimax-h3",
      "published_at": "2026-09-08",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "fal.ai MiniMax H3 Max & Turbo Video Models Live with Dynamic Multimodal Dispatch",
        "zh": "fal.ai MiniMax H3 Max 与 Turbo 视频模型上线，支持动态多模态分发"
      },
      "summary": {
        "en": "MiniMax H3 Max and H3 Max Turbo (<3s rendering) video models are live on OmniMux via fal.ai, supporting dynamic multimodal routing across text-to-video, image-to-video, and reference-to-video.",
        "zh": "基于 fal.ai 后训练的 MiniMax H3 Max 旗舰版与 H3 Max Turbo 极速版（约3秒出片）视频模型已在 OmniMux 上线，支持文生视频、图生视频及参考生视频的动态多模态参数自适应分发。"
      },
      "models": [
        "minimax/h3-max",
        "minimax/h3-max-turbo"
      ],
      "tags": [
        "video",
        "multimodal",
        "fal.ai",
        "minimax"
      ],
      "links": [
        {
          "label": {
            "en": "MiniMax H3 Max",
            "zh": "MiniMax H3 Max"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/minimax-h3-max",
            "zh": "/zh/api-reference/video-series/models/minimax-h3-max"
          }
        },
        {
          "label": {
            "en": "MiniMax H3 Max Turbo",
            "zh": "MiniMax H3 Max Turbo"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/minimax-h3-max-turbo",
            "zh": "/zh/api-reference/video-series/models/minimax-h3-max-turbo"
          }
        },
        {
          "label": {
            "en": "Pricing",
            "zh": "定价"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        }
      ],
      "body": {
        "en": "## fal.ai MiniMax H3 Max & Turbo\n\n**MiniMax H3 Max** and **MiniMax H3 Max Turbo** video generation models are now live on OmniMux.\n\n### Highlights\n\n- **Dynamic Multimodal Dispatch**: Clients only need to configure two base models (`minimax/h3-max` and `minimax/h3-max-turbo`). Passing prompt-only routes to text-to-video, adding `image_url` routes to image-to-video, and passing `video_url` routes to reference-to-video automatically.\n- **Ultra-Fast Turbo Performance**: Sub-3-second end-to-end rendering for 5-second video clips.\n- **Two-Phase Resilient Polling**: Transparent inline result resolution with CDN 404 lag tolerance.\n- **Standard API**: Create via `POST /v1/video/generations` and query via `GET /v1/tasks/{task_id}`.",
        "zh": "## fal.ai MiniMax H3 Max 与 Turbo 视频模型\n\n**MiniMax H3 Max** 旗舰版与 **MiniMax H3 Max Turbo** 极速版视频生成模型现已在 OmniMux 正式上线。\n\n### 核心特性\n\n- **动态多模态分发**：客户端仅需配置两个基础模型（`minimax/h3-max` 和 `minimax/h3-max-turbo`）。只传文字走文生视频，传入 `image_url` 自动切图生视频，传入 `video_url` 自动切参考生视频，心智极简。\n- **Turbo 极速出片**：5 秒视频片段端到端约 3 秒即可出片（模型推理仅约 1.5 秒）。\n- **双阶段内聚轮询与容错**：成片 URL 就地解析，并具备针对 CDN 延迟的自愈防误杀能力。\n- **标准协议**：支持通过 `POST /v1/video/generations` 异步提交任务，并通过 `GET /v1/tasks/{task_id}` 轮询结果。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-07-social-provider-required",
      "published_at": "2026-09-07",
      "type": "breaking",
      "modality": [
        "publishing"
      ],
      "title": {
        "en": "Social operations require an explicit provider",
        "zh": "社媒操作必须明确指定账号来源"
      },
      "summary": {
        "en": "Pass tiktok_direct or zernio on five account and publishing operations. Upgrade the CLI and update saved commands.",
        "zh": "五项账号及发布操作必须传入 tiktok_direct 或 zernio；请升级 CLI 并更新已有命令。"
      },
      "models": [],
      "tags": [
        "provider",
        "CLI",
        "migration"
      ],
      "links": [
        {
          "label": {
            "en": "List accounts",
            "zh": "账号列表"
          },
          "href": {
            "en": "/en/api-reference/publishing/list-accounts",
            "zh": "/zh/api-reference/publishing/list-accounts"
          }
        }
      ],
      "body": {
        "en": "The gateway began requiring `provider` at **2026-09-07 07:25:20 UTC+8 (2026-09-06T23:25:20Z)** for listing accounts, starting a connection, disconnecting, creating a post and reading post status. Use `tiktok_direct` for official TikTok accounts or `zernio` for Zernio accounts. GET/DELETE carry it in the query; POST carries it in JSON. Missing or unknown values return `400 invalid-provider`; a resource from another source returns `409 account-provider-mismatch` or `409 post-provider-mismatch`. There is no default source.\n\n### CLI migration\n\nUse CLI **0.4.0**: `omnimux update` for binaries, or `npm install -g @omnimux/cli@0.4.0` for npm. Update all saved commands to include `--provider`, for example `omnimux social accounts --provider zernio`. For a read without upstream refresh, use `omnimux social post-status <post-id> --provider zernio --no-refresh`. Updating a binary does not update your scripts or prove all clients have upgraded.\n\nCLI 0.4.0 also includes the previously merged hosted commands and video skill updates. Credentials use a mode-0600 file on every platform; Keychain-only installations need credentials configured securely. The Agent RPC envelope remains `contract_version: 1`.\n\nCLI 0.4.0 was published on **2026-09-07 at 01:47:36 UTC (09:47:36 UTC+8)**. Both the public binary release and npm package are version 0.4.0.\n\nStandalone binaries do not include skill files. Install the npm package to use or update the bundled skills; a binary-only update can report that the skill tree is missing even when the CLI upgrade succeeds.",
        "zh": "网关从 **2026-09-07 07:25:20 UTC+8（2026-09-06T23:25:20Z）** 起，要求账号列表、发起连接、解绑、创建帖子及查询帖子状态五项操作显式传入 `provider`。官方 TikTok 账号使用 `tiktok_direct`，Zernio 账号使用 `zernio`。GET/DELETE 使用查询参数，POST 使用 JSON 字段。缺失或未知值返回 `400 invalid-provider`；跨来源资源返回 `409 account-provider-mismatch` 或 `409 post-provider-mismatch`。没有默认来源。\n\n### CLI 迁移\n\n使用 CLI **0.4.0**：二进制安装执行 `omnimux update`，npm 安装执行 `npm install -g @omnimux/cli@0.4.0`。同步修改已有命令，补充 `--provider`，例如 `omnimux social accounts --provider zernio`。不刷新上游的状态读取示例：`omnimux social post-status <post-id> --provider zernio --no-refresh`。升级二进制不会修改脚本，也不代表全部客户端都已升级。\n\nCLI 0.4.0 同时包含此前合入的 hosted 命令和视频技能更新。各平台凭据统一保存在权限为 0600 的文件中；仅使用 Keychain 的旧安装需要安全配置凭据。Agent RPC 仍为 `contract_version: 1`。\n\nCLI 0.4.0 于 **2026-09-07 01:47:36 UTC（09:47:36 UTC+8）** 公开发布。公开二进制与 npm 包版本均为 0.4.0。\n\n独立二进制不包含技能文件。需要使用或更新随包技能时，请安装 npm 包；仅升级二进制时，即使 CLI 升级成功，技能同步仍可能报告缺少技能目录。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-09-07-account-quota-errors",
      "published_at": "2026-09-07",
      "type": "platform",
      "modality": [
        "text"
      ],
      "title": {
        "en": "Account quota errors on OpenAI text endpoints",
        "zh": "OpenAI 文本接口统一账户额度错误"
      },
      "summary": {
        "en": "Local account-quota failures return 402 insufficient_quota, preventing authentication misclassification and unnecessary retries.",
        "zh": "本地账户额度不足返回 402 insufficient_quota，避免误判为密钥失效或触发无效重试。"
      },
      "models": [],
      "tags": [
        "errors",
        "quota",
        "compatibility"
      ],
      "links": [
        {
          "label": {
            "en": "Connection and usage",
            "zh": "连接与使用"
          },
          "href": {
            "en": "/en/faqs/connection-usage",
            "zh": "/zh/faqs/connection-usage"
          }
        }
      ],
      "body": {
        "en": "OpenAI Completions, Chat Completions, Responses, and Responses Compact return HTTP **402** with `error.type` and `error.code` set to `insufficient_quota` when the local account balance or subscription allowance cannot cover pre-consume. Streaming requests receive the JSON error before SSE starts. Check balance or allowance before trying again; do not treat this as invalid credentials or temporary rate limiting. Token errors, upstream provider errors, other protocols, actual 401/403, ordinary 429, and 5xx behavior are unchanged.\n\nGateway enforcement began on September 7, 2026 at 07:25:20 UTC+8 (2026-09-06T23:25:20Z).",
        "zh": "OpenAI Completions、Chat Completions、Responses 和 Responses Compact 在本地账户余额或订阅额度不足以完成预扣时，返回 HTTP **402**，`error.type` 与 `error.code` 均为 `insufficient_quota`。流式请求在 SSE 建立前返回 JSON 错误。应先检查余额或额度，再决定是否重新请求；不要将其当作密钥失效或临时限流。token 错误、上游渠道错误、其他协议，以及真正的 401/403、普通 429 和 5xx 行为保持不变。\n\n网关于 2026 年 9 月 7 日 07:25:20 UTC+8（2026-09-06T23:25:20Z）上线该响应。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-08-31-catalog-sync",
      "published_at": "2026-08-31",
      "type": "model_launch",
      "modality": [
        "text",
        "image",
        "video",
        "audio"
      ],
      "title": {
        "en": "Catalog synchronization: Gemini 3.7, GLM 5.3, Wan 3.0, Seedance 2.0, Grok Imagine, Audio TTS & STT",
        "zh": "模型目录全量同步：Gemini 3.7、GLM 5.3、Wan 3.0、Seedance 2.0、Grok Imagine、音频与语音转录"
      },
      "summary": {
        "en": "Comprehensive documentation alignment across all 108 live models in the OmniMux catalog.",
        "zh": "全面对齐 OmniMux 线上 108 个在售模型的开发文档与 API 参考手册。"
      },
      "models": [
        "gemini-3.7-flash",
        "glm-5.3",
        "grok-4.6",
        "deepseek-v4-flash-vision-exp",
        "wan-3.0",
        "seedance-2-0",
        "gpt-4o-mini-tts",
        "whisper-1"
      ],
      "tags": [
        "text",
        "image",
        "video",
        "audio"
      ],
      "links": [
        {
          "label": {
            "en": "Pricing",
            "zh": "定价"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        },
        {
          "label": {
            "en": "Audio Speech",
            "zh": "语音合成"
          },
          "href": {
            "en": "/en/api-reference/audio-series/models/gpt-4o-mini-tts",
            "zh": "/zh/api-reference/audio-series/models/gpt-4o-mini-tts"
          }
        }
      ],
      "body": {
        "en": "## Language series\n\n- **Gemini:** Added `gemini-3.7-flash`\n- **GLM:** Added `glm-5.3`\n- **Grok:** Added `grok-4.6`\n- **DeepSeek:** Added `deepseek-v4-flash-vision-exp`\n\n## Image & Video series\n\n- Canonical model ID hyphenation and full coverage for **Wan 3.0** (`wan-3.0`, `wan-3.0-prime`, `wan-3.0-prime-ref`, `wan-3.0-ref`), **Seedance 2.0** (`seedance-2-0`, `seedance-2-0-fast`, `seedance-2-0-mini`, `seedance-2-5`), **Kling** (`kling-v3`, `kling-v2-6`, `kling-avatar`, `kling-o3`, `kling-v3-motion-control`), **Grok Imagine**, and **Midjourney**.\n- Retired legacy Omni Flash duration-fixed SKUs.\n\n## Audio series\n\n- `gpt-4o-mini-tts` (`POST /v1/audio/speech`)\n- `whisper-1` (`POST /v1/audio/transcriptions`)\n- `suno` (`POST /v1/video/generations`)",
        "zh": "## 语言系列\n\n- **Gemini：** 增加 `gemini-3.7-flash`\n- **GLM：** 增加 `glm-5.3`\n- **Grok：** 增加 `grok-4.6`\n- **DeepSeek：** 增加 `deepseek-v4-flash-vision-exp`\n\n## 图像与视频系列\n\n- 全量补齐 **Wan 3.0**（`wan-3.0`、`wan-3.0-prime`、`wan-3.0-prime-ref`、`wan-3.0-ref`）、**Seedance 2.0**（`seedance-2-0`、`seedance-2-0-fast`、`seedance-2-0-mini`、`seedance-2-5`）、**Kling**（`kling-v3`、`kling-v2-6`、`kling-avatar`、`kling-o3`、`kling-v3-motion-control`）、**Grok Imagine** 与 **Midjourney** 文档。\n- 下架并清理已停售的 Omni Flash 固定时长 SKU。\n\n## 音频系列\n\n- `gpt-4o-mini-tts`（`POST /v1/audio/speech` 语音合成）\n- `whisper-1`（`POST /v1/audio/transcriptions` 语音转录）\n- `suno`（`POST /v1/video/generations` 音乐生成）"
      },
      "schema_version": 1
    },
    {
      "id": "2026-08-31-media-catalog",
      "published_at": "2026-08-31",
      "type": "model_launch",
      "modality": [
        "image",
        "video"
      ],
      "title": {
        "en": "Seedream 5.0 Pro, Qwen Image 3.0, Grok Imagine Image 2.0, Kling O3, Omni 1.1, PixVerse v6, Vidu Q3",
        "zh": "Seedream 5.0 Pro、Qwen Image 3.0、Grok Imagine Image 2.0、Kling O3、Omni 1.1、PixVerse v6、Vidu Q3"
      },
      "summary": {
        "en": "Seven image and video models are live on OmniMux.",
        "zh": "七个图像 / 视频模型已在 OmniMux 可调用。"
      },
      "models": [
        "seedream-5-0-pro",
        "qwen-image-3-0",
        "grok-imagine-image-2-0",
        "kling-o3",
        "gemini-omni-1.1",
        "pixverse-v6",
        "vidu-q3"
      ],
      "tags": [
        "image",
        "video"
      ],
      "links": [
        {
          "label": {
            "en": "Pricing",
            "zh": "定价"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        },
        {
          "label": {
            "en": "Video task poll",
            "zh": "视频任务查询"
          },
          "href": {
            "en": "/en/api-reference/tasks/video-task",
            "zh": "/zh/api-reference/tasks/video-task"
          }
        }
      ],
      "body": {
        "en": "## Image\n\n- **Model IDs:** `seedream-5-0-pro`, `qwen-image-3-0`, `grok-imagine-image-2-0`\n- **API:** `POST /v1/images/generations`\n\n## Video\n\n- **Model IDs:** `kling-o3`, `gemini-omni-1.1`, `pixverse-v6`, `vidu-q3`\n- **API:** async `POST` / `GET` `/v1/video/generations`\n\n`kling-o3` is a new id and does not replace `kling-v3`. `gemini-omni-1.1` is distinct from `gemini-omni-flash`. Pin IDs from live pricing.",
        "zh": "## 图像\n\n- **模型 ID：** `seedream-5-0-pro`、`qwen-image-3-0`、`grok-imagine-image-2-0`\n- **API：** `POST /v1/images/generations`\n\n## 视频\n\n- **模型 ID：** `kling-o3`、`gemini-omni-1.1`、`pixverse-v6`、`vidu-q3`\n- **API：** 异步 `POST` / `GET` `/v1/video/generations`\n\n`kling-o3` 是新 ID，不覆盖 `kling-v3`。`gemini-omni-1.1` 与 `gemini-omni-flash` 不同。精确 ID 以线上定价为准。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-08-25-social-analytics-reader-seats",
      "published_at": "2026-08-25",
      "type": "platform",
      "modality": [
        "platform",
        "text"
      ],
      "title": {
        "en": "Social analytics & inbox suite, Jina Reader extractor, and social seats",
        "zh": "上线社媒分析与私信分析套件、Jina Reader 网页提取及社媒席位管理"
      },
      "summary": {
        "en": "Landed 11 social analytics and inbox endpoints under `/api/social/v1`, Jina Reader `/v1/reader` gateway relay with output-token billing, and social seats status API.",
        "zh": "上线 `/api/social/v1` 11 个社媒分析与私信分析端点、Jina Reader `/v1/reader` 网页提取中继（按输出 Token 计费）以及社媒席位管理接口。"
      },
      "models": [
        "jina-reader-v1"
      ],
      "tags": [
        "publishing",
        "analytics",
        "reader",
        "social"
      ],
      "links": [
        {
          "label": {
            "en": "Jina Reader Extractor",
            "zh": "Jina Reader 网页提取"
          },
          "href": {
            "en": "/en/api-reference/reader-series/models/jina-reader-v1",
            "zh": "/zh/api-reference/reader-series/models/jina-reader-v1"
          }
        },
        {
          "label": {
            "en": "Daily Metrics",
            "zh": "大盘时序指标"
          },
          "href": {
            "en": "/en/api-reference/publishing/analytics/daily-metrics",
            "zh": "/zh/api-reference/publishing/analytics/daily-metrics"
          }
        },
        {
          "label": {
            "en": "Social Seats",
            "zh": "查询席位"
          },
          "href": {
            "en": "/en/api-reference/publishing/seats",
            "zh": "/zh/api-reference/publishing/seats"
          }
        }
      ],
      "body": {
        "en": "## Social Analytics & Inbox Suite\n\nOmniMux now exposes 11 RESTful analytics endpoints under `/api/social/v1` for comprehensive audience and engagement intelligence:\n\n- **Analytics:** Daily metrics timeseries, 7x24 weekly best time to post heatmap, posting frequency cadence model, content decay curve, follower growth snapshots, posts performance ranking, and external posts incremental synchronization.\n- **Inbox Analytics:** Message volume trends, response time (TTR) distribution, 7x24 incoming activity heatmap, and source breakdown across connected platforms.\n\n## Jina Reader Relay (`POST /v1/reader`)\n\n- **Model ID:** `jina-reader-v1`\n- **Capability:** URL to clean Markdown / structured JSON text extraction with CSS selectors filtering and output-token billing.\n- **Auth:** Standard gateway Bearer `sk-...` token.\n\n## Social Seats Management (`GET /api/social/v1/seats`)\n\n- Inspects current active connected social accounts, total quota capacity, and vacant seat availability.",
        "zh": "## 社媒分析与私信分析套件\n\nOmniMux 在 `/api/social/v1` 路由组下全新接入 11 个 RESTful 深度数据分析与私信洞察端点：\n\n- **社媒分析（Analytics）：** 宏观大盘时序指标、7×24小时最佳发布时间热力矩阵、发帖频次与互动率相关性模型、内容生命周期衰减曲线、多平台粉丝增长快照、单帖与列表深度指标排行、外部平台存量帖子增量同步。\n- **私信分析（Inbox Analytics）：** 私信与互动消息量时序走势、响应时效（TTR）耗时分布、7×24小时私信时段热力图、跨平台私信渠道来源占比。\n\n## Jina Reader 网页提取中继（`POST /v1/reader`）\n\n- **模型 ID：** `jina-reader-v1`\n- **能力：** 将网页 URL 高效转换为大模型友好的 Markdown 或结构化 JSON 文本，支持 CSS 选择器过滤、图片摘要与按输出 Token 结算。\n- **鉴权：** 标准网关 Bearer `sk-...` 凭据。\n\n## 社媒席位管理（`GET /api/social/v1/seats`）\n\n- 查询当前用户绑定的社媒账号数量、席位配额容量与可用空坑状态。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-08-16-index-tts-ltx-docs",
      "published_at": "2026-08-16",
      "type": "model_launch",
      "modality": [
        "audio",
        "video"
      ],
      "title": {
        "en": "Index TTS voice clone plus LTX digital-human contract",
        "zh": "上架 Index TTS 声音克隆，并更正 LTX 数字人合同"
      },
      "summary": {
        "en": "`index-tts` is live for GxgenAI voice clone. `ltx2.3-kj` docs now describe required photo + audio nodeInfoList, not text-to-video.",
        "zh": "`index-tts` 已上架，用于 GxgenAI 声音克隆。`ltx2.3-kj` 文档改为必填人设图+音频 nodeInfoList，不再写成文生视频。"
      },
      "models": [
        "index-tts",
        "ltx-2-3-kj"
      ],
      "tags": [
        "audio",
        "video",
        "gxgenai"
      ],
      "links": [
        {
          "label": {
            "en": "Index TTS",
            "zh": "Index TTS"
          },
          "href": {
            "en": "/en/api-reference/audio-series/models/index-tts",
            "zh": "/zh/api-reference/audio-series/models/index-tts"
          }
        },
        {
          "label": {
            "en": "LTX digital human",
            "zh": "LTX 数字人对口型"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/ltx-2-3-kj",
            "zh": "/zh/api-reference/video-series/models/ltx-2-3-kj"
          }
        }
      ],
      "body": {
        "en": "## New model\n\n- **Model ID:** `index-tts`\n- **Path:** `POST` / `GET` `/v1/video/generations` (GxgenAI task path; result is audio)\n- **Contract:** `metadata.nodeInfoList` required — node `4/audio` (reference clip) and `7/text` (script)\n\n## Docs fix\n\n- **`ltx-2-3-kj`** is a photo + audio lip-sync model. Prompt-only text-to-video examples were incorrect. Send image node `444` and audio node `1755`.",
        "zh": "## 新模型\n\n- **模型 ID：** `index-tts`\n- **路径：** `POST` / `GET` `/v1/video/generations`（GxgenAI 任务路径，产物为音频）\n- **合同：** 必须传 `metadata.nodeInfoList` — 节点 `4/audio`（参考音）和 `7/text`（文稿）\n\n## 文档更正\n\n- **`ltx-2-3-kj`** 是人设静帧 + 音频对口型，不是文生视频。请传图片节点 `444` 和音频节点 `1755`。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-08-09-platform-api-updates",
      "published_at": "2026-08-09",
      "type": "platform",
      "modality": [
        "platform"
      ],
      "title": {
        "en": "API Updates timeline opens",
        "zh": "API 更新时间线上线"
      },
      "summary": {
        "en": "Public developer changelog for model launches and platform changes.",
        "zh": "面向开发者的公开更新时间线：模型上新与平台变更。"
      },
      "models": [],
      "tags": [
        "platform"
      ],
      "links": [
        {
          "label": {
            "en": "Quickstart",
            "zh": "快速开始"
          },
          "href": {
            "en": "/en/quickstart",
            "zh": "/zh/quickstart"
          }
        },
        {
          "label": {
            "en": "Pricing",
            "zh": "定价与账户"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        }
      ],
      "body": {
        "en": "## API Updates\n\nOmniMux now publishes a dated **API Updates** feed so you can track gateway model launches and platform changes in one place.\n\n### What we log\n\n- **New models** — callable model IDs on live pricing\n- **Model updates** — modes, durations, resolutions, parameters\n- **Pricing** — material rate changes\n- **Platform** — auth, domains, public HTTP surfaces\n\n### How to call OmniMux\n\n| Surface | URL |\n| --- | --- |\n| Console | `https://omnimux.ai` |\n| OpenAI-compatible API | `https://api.omnimux.ai/v1` |\n| Docs | `https://docs.omnimux.ai` |\n| Status | `https://status.omnimux.ai` |\n\nAuth for relay: Bearer `sk-…` from the console. Full catalog always lives on console pricing — this page is the **change timeline**.",
        "zh": "## API 更新\n\nOmniMux 现提供按日期排列的 **API 更新** 时间线，便于跟踪网关模型上新与平台变更。\n\n### 记录范围\n\n- **新模型** — 线上定价可调用的模型 ID\n- **模型更新** — 模式、时长、分辨率、参数\n- **价格调整** — 实质费率变化\n- **平台** — 鉴权、域名、公开 HTTP 面\n\n### 使用入口\n\n| 面 | URL |\n| --- | --- |\n| 控制台 | `https://omnimux.ai` |\n| OpenAI 兼容 API | `https://api.omnimux.ai/v1` |\n| 文档 | `https://docs.omnimux.ai` |\n| 状态页 | `https://status.omnimux.ai` |\n\nRelay 鉴权：控制台签发的 Bearer `sk-…`。完整目录以控制台定价为准——本页是 **变更时间线**。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-08-09-social-data",
      "published_at": "2026-08-09",
      "type": "model_launch",
      "modality": [
        "social-data"
      ],
      "title": {
        "en": "Social data models — TikTok, Instagram, X, YouTube",
        "zh": "社交数据模型 — TikTok / Instagram / X / YouTube"
      },
      "summary": {
        "en": "Read-only social data via Chat Completions-style model IDs.",
        "zh": "只读社交数据：Chat Completions 形态的模型 ID。"
      },
      "models": [
        "tiktok-user",
        "tiktok-posts",
        "tiktok-video",
        "tiktok-search",
        "instagram-user",
        "instagram-posts",
        "instagram-post",
        "instagram-search",
        "x-user",
        "x-posts",
        "x-tweet",
        "x-search",
        "youtube-user",
        "youtube-posts",
        "youtube-video",
        "youtube-search"
      ],
      "tags": [
        "social-data"
      ],
      "links": [
        {
          "label": {
            "en": "TikTok user profile",
            "zh": "TikTok 用户资料"
          },
          "href": {
            "en": "/en/api-reference/social-data/tiktok/user-profile",
            "zh": "/zh/api-reference/social-data/tiktok/user-profile"
          }
        },
        {
          "label": {
            "en": "X tweet detail",
            "zh": "X 推文详情"
          },
          "href": {
            "en": "/en/api-reference/social-data/x/tweet-detail",
            "zh": "/zh/api-reference/social-data/x/tweet-detail"
          }
        },
        {
          "label": {
            "en": "Create post (publishing)",
            "zh": "创建帖子（社媒发布）"
          },
          "href": {
            "en": "/en/api-reference/publishing/create-post",
            "zh": "/zh/api-reference/publishing/create-post"
          }
        }
      ],
      "body": {
        "en": "## Social data (read-only)\n\n**Social data** model IDs are available on OmniMux for read-only public profile and content style access. This is **not** social publishing.\n\n### Model IDs\n\n| Platform | IDs |\n| --- | --- |\n| TikTok | `tiktok-user`, `tiktok-posts`, `tiktok-video`, `tiktok-search` |\n| Instagram | `instagram-user`, `instagram-posts`, `instagram-post`, `instagram-search` |\n| X | `x-user`, `x-posts`, `x-tweet`, `x-search` |\n| YouTube | `youtube-user`, `youtube-posts`, `youtube-video`, `youtube-search` |\n\n### Publishing (different surface)\n\nConnect accounts and create posts use **`/api/social/v1/*`** with access token + `New-Api-User`. Do not mix that auth with gateway `sk-` Chat Completions.",
        "zh": "## 社交数据（只读）\n\n**社交数据** 模型 ID 已在 OmniMux 可调用，用于只读的公开资料/内容类访问。**不是** 社媒发布。\n\n### 模型 ID\n\n| 平台 | ID |\n| --- | --- |\n| TikTok | `tiktok-user`、`tiktok-posts`、`tiktok-video`、`tiktok-search` |\n| Instagram | `instagram-user`、`instagram-posts`、`instagram-post`、`instagram-search` |\n| X | `x-user`、`x-posts`、`x-tweet`、`x-search` |\n| YouTube | `youtube-user`、`youtube-posts`、`youtube-video`、`youtube-search` |\n\n### 社媒发布（另一套面）\n\n连接账户与发帖走 **`/api/social/v1/*`**，鉴权为 access token + `New-Api-User`。勿与网关 `sk-` Chat Completions 混用。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-08-07-seedance-2-5",
      "published_at": "2026-08-07",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "Seedance 2.5 — longer video with multimodal references",
        "zh": "Seedance 2.5 — 更长视频与多模态参考"
      },
      "summary": {
        "en": "Seedance 2.5 video generation is available on OmniMux.",
        "zh": "Seedance 2.5 视频生成已在 OmniMux 可调用。"
      },
      "models": [
        "seedance-2-5"
      ],
      "tags": [
        "video"
      ],
      "links": [
        {
          "label": {
            "en": "Video task poll",
            "zh": "视频任务查询"
          },
          "href": {
            "en": "/en/api-reference/tasks/video-task",
            "zh": "/zh/api-reference/tasks/video-task"
          }
        },
        {
          "label": {
            "en": "Pricing",
            "zh": "定价"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        }
      ],
      "body": {
        "en": "## Seedance 2.5\n\n**Seedance 2.5** is now available on OmniMux for async video generation.\n\n### New model\n\n- **Model ID:** `seedance-2-5`\n- **Create:** `POST /v1/video/generations`\n- **Poll:** `GET /v1/video/generations/{task_id}`\n- **Highlights:** longer single-pass generation and multimodal reference inputs (text / image / video / audio style workflows)\n\nPin the exact ID from live pricing before integrating.",
        "zh": "## Seedance 2.5\n\n**Seedance 2.5** 现已在 OmniMux 可调用（异步视频生成）。\n\n### 新模型\n\n- **模型 ID：** `seedance-2-5`\n- **创建：** `POST /v1/video/generations`\n- **查询：** `GET /v1/video/generations/{task_id}`\n- **要点：** 更长单次生成，支持多模态参考输入（文本 / 图像 / 视频 / 音频类工作流）\n\n接入前请以线上定价的精确 ID 为准。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-07-30-minimax-h3",
      "published_at": "2026-07-30",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "MiniMax H3 — text-to-video and frame variants",
        "zh": "MiniMax H3 — 文生视频与帧变体"
      },
      "summary": {
        "en": "MiniMax H3 video routes are live on OmniMux.",
        "zh": "MiniMax H3 视频路由已在 OmniMux 上线。"
      },
      "models": [
        "minimax-h3",
        "minimax-h3-t2v",
        "minimax-h3-flf",
        "minimax-h3-fl2va",
        "minimax-h3-endframe"
      ],
      "tags": [
        "video"
      ],
      "links": [
        {
          "label": {
            "en": "MiniMax H3 Video",
            "zh": "MiniMax H3 视频"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/minimax-h3",
            "zh": "/zh/api-reference/video-series/models/minimax-h3"
          }
        },
        {
          "label": {
            "en": "MiniMax H3 T2V",
            "zh": "MiniMax H3 文生视频"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/minimax-h3-t2v",
            "zh": "/zh/api-reference/video-series/models/minimax-h3-t2v"
          }
        },
        {
          "label": {
            "en": "Video task poll",
            "zh": "视频任务查询"
          },
          "href": {
            "en": "/en/api-reference/tasks/video-task",
            "zh": "/zh/api-reference/tasks/video-task"
          }
        }
      ],
      "body": {
        "en": "## MiniMax H3\n\n**MiniMax H3** video models are now available on OmniMux.\n\n### New models\n\n| Model ID | Role |\n| --- | --- |\n| `minimax-h3` | Standard text/image-to-video |\n| `minimax-h3-t2v` | Text-to-video |\n| `minimax-h3-flf` | First/last-frame style I2V |\n| `minimax-h3-fl2va` | FL2VA-oriented path |\n| `minimax-h3-endframe` | End-frame reference I2V |\n\n### API\n\n- **Create:** `POST /v1/video/generations`\n- **Poll:** `GET /v1/video/generations/{task_id}`\n\nAsync only — submit once, then poll task status.",
        "zh": "## MiniMax H3\n\n**MiniMax H3** 视频模型现已在 OmniMux 可调用。\n\n### 新模型\n\n| 模型 ID | 角色 |\n| --- | --- |\n| `minimax-h3` | 标准文/图生视频统一路由 |\n| `minimax-h3-t2v` | 文生视频 |\n| `minimax-h3-flf` | 首末帧类 I2V |\n| `minimax-h3-fl2va` | FL2VA 路径 |\n| `minimax-h3-endframe` | 尾帧参考 I2V |\n\n### 调用\n\n- **创建：** `POST /v1/video/generations`\n- **查询：** `GET /v1/video/generations/{task_id}`\n\n异步任务：提交后轮询状态。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-07-28-grok-imagine-video-1-5",
      "published_at": "2026-07-28",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "Grok Imagine Video 1.5",
        "zh": "Grok Imagine Video 1.5"
      },
      "summary": {
        "en": "Image-to-video with Grok Imagine Video 1.5 on OmniMux.",
        "zh": "Grok Imagine Video 1.5 图生视频已在 OmniMux 可调用。"
      },
      "models": [
        "grok-imagine-video-1-5"
      ],
      "tags": [
        "video"
      ],
      "links": [
        {
          "label": {
            "en": "Grok Imagine Video 1.5",
            "zh": "Grok Imagine Video 1.5"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/grok-imagine-video-1-5",
            "zh": "/zh/api-reference/video-series/models/grok-imagine-video-1-5"
          }
        },
        {
          "label": {
            "en": "Video task poll",
            "zh": "视频任务查询"
          },
          "href": {
            "en": "/en/api-reference/tasks/video-task",
            "zh": "/zh/api-reference/tasks/video-task"
          }
        }
      ],
      "body": {
        "en": "## Grok Imagine Video 1.5\n\n**Grok Imagine Video 1.5** is now available on OmniMux.\n\n### New model\n\n- **Model ID:** `grok-imagine-video-1-5`\n- **Mode:** image-to-video style generation via async video API\n- **Create / poll:** `POST` / `GET` `/v1/video/generations`\n\nSee the model page for parameters and the task poll contract.",
        "zh": "## Grok Imagine Video 1.5\n\n**Grok Imagine Video 1.5** 现已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `grok-imagine-video-1-5`\n- **模式：** 异步视频 API 下的图生视频类能力\n- **创建 / 查询：** `POST` / `GET` `/v1/video/generations`\n\n参数与任务查询约定见模型页与任务文档。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-07-25-claude-opus-5",
      "published_at": "2026-07-25",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "Claude Opus 5",
        "zh": "Claude Opus 5"
      },
      "summary": {
        "en": "Claude Opus 5 is available via Chat Completions on OmniMux.",
        "zh": "Claude Opus 5 已通过 Chat Completions 在 OmniMux 可调用。"
      },
      "models": [
        "claude-opus-5"
      ],
      "tags": [
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "Claude complete",
            "zh": "Claude 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/claude/complete",
            "zh": "/zh/api-reference/text-series/claude/complete"
          }
        }
      ],
      "body": {
        "en": "## Claude Opus 5\n\n**Claude Opus 5** is now available on OmniMux through OpenAI-compatible Chat Completions.\n\n### New model\n\n- **Model ID:** `claude-opus-5`\n- **Endpoint:** `https://api.omnimux.ai/v1` (Chat Completions)\n- **Docs:** one complete page per brand — `model` is an enum on the Claude complete page\n\nRelated Claude IDs on the gateway include other Opus / Sonnet / Haiku SKUs on live pricing.",
        "zh": "## Claude Opus 5\n\n**Claude Opus 5** 现已通过 OpenAI 兼容 Chat Completions 在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `claude-opus-5`\n- **端点：** `https://api.omnimux.ai/v1`（Chat Completions）\n- **文档：** 每个品牌一页完整参数 — `model` 为 Claude 完整参数页内枚举\n\n网关上同族还有其他 Opus / Sonnet / Haiku SKU，以线上定价为准。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-07-22-gemini-3-6-flash",
      "published_at": "2026-07-22",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "Gemini 3.6 Flash and Gemini 3.5 Flash",
        "zh": "Gemini 3.6 Flash 与 Gemini 3.5 Flash"
      },
      "summary": {
        "en": "Gemini Flash-class language models on OmniMux.",
        "zh": "Gemini Flash 档语言模型已在 OmniMux 可调用。"
      },
      "models": [
        "gemini-3.6-flash",
        "gemini-3.5-flash"
      ],
      "tags": [
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "Gemini complete",
            "zh": "Gemini 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/gemini/complete",
            "zh": "/zh/api-reference/text-series/gemini/complete"
          }
        }
      ],
      "body": {
        "en": "## Gemini Flash models\n\nNew Gemini Flash-class models are available on OmniMux.\n\n### New models\n\n- **Model IDs:** `gemini-3.6-flash`, `gemini-3.5-flash`\n- **API:** OpenAI-compatible Chat Completions at `https://api.omnimux.ai/v1`\n\nAdditional Gemini IDs (Pro / preview / lite) may also appear on live pricing — pin the exact ID from the console or Gemini complete page.",
        "zh": "## Gemini Flash 模型\n\nGemini Flash 档模型现已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `gemini-3.6-flash`、`gemini-3.5-flash`\n- **API：** OpenAI 兼容 Chat Completions：`https://api.omnimux.ai/v1`\n\n线上定价中可能还有 Pro / preview / lite 等 Gemini ID——请以控制台或 Gemini 完整参数页为准。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-07-17-kimi-k3",
      "published_at": "2026-07-17",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "Kimi K3 — long-context reasoning",
        "zh": "Kimi K3 — 长上下文推理"
      },
      "summary": {
        "en": "Kimi K3 is available on OmniMux Chat Completions.",
        "zh": "Kimi K3 已在 OmniMux Chat Completions 可调用。"
      },
      "models": [
        "kimi-k3"
      ],
      "tags": [
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "Kimi complete",
            "zh": "Kimi 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/kimi/complete",
            "zh": "/zh/api-reference/text-series/kimi/complete"
          }
        }
      ],
      "body": {
        "en": "## Kimi K3\n\n**Kimi K3** is now available on OmniMux.\n\n### New model\n\n- **Model ID:** `kimi-k3`\n- **API:** Chat Completions at `https://api.omnimux.ai/v1`\n- **Focus:** long-context reasoning and agent-style workflows\n\nSee the Kimi complete page for the full model enum.",
        "zh": "## Kimi K3\n\n**Kimi K3** 现已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `kimi-k3`\n- **API：** Chat Completions：`https://api.omnimux.ai/v1`\n- **定位：** 长上下文推理与 Agent 类工作流\n\n完整模型枚举见 Kimi 完整参数页。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-07-16-suno",
      "published_at": "2026-07-16",
      "type": "model_launch",
      "modality": [
        "audio"
      ],
      "title": {
        "en": "Suno — music generation",
        "zh": "Suno — 音乐生成"
      },
      "summary": {
        "en": "Suno music generation is available on OmniMux.",
        "zh": "Suno 音乐生成已在 OmniMux 可调用。"
      },
      "models": [
        "suno"
      ],
      "tags": [
        "audio"
      ],
      "links": [
        {
          "label": {
            "en": "Pricing",
            "zh": "定价"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        }
      ],
      "body": {
        "en": "## Suno\n\n**Suno** music generation is now available on OmniMux.\n\n### New model\n\n- **Model ID:** `suno`\n- **Use:** AI music generation workflows via the gateway catalog\n\nConfirm billing and request shape from live pricing / console before production traffic.",
        "zh": "## Suno\n\n**Suno** 音乐生成现已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `suno`\n- **用途：** 通过网关目录进行 AI 音乐生成\n\n上线生产前请在控制台定价中确认计费与请求形态。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-07-01-nano-banana",
      "published_at": "2026-07-01",
      "type": "model_launch",
      "modality": [
        "image"
      ],
      "title": {
        "en": "Nano Banana 2 and Nano Banana Pro",
        "zh": "Nano Banana 2 与 Nano Banana Pro"
      },
      "summary": {
        "en": "Nano Banana image models on OmniMux.",
        "zh": "Nano Banana 图像模型已在 OmniMux 可调用。"
      },
      "models": [
        "nano-banana-2",
        "nano-banana-pro"
      ],
      "tags": [
        "image"
      ],
      "links": [
        {
          "label": {
            "en": "Nano Banana 2",
            "zh": "Nano Banana 2"
          },
          "href": {
            "en": "/en/api-reference/image-series/models/nano-banana-2",
            "zh": "/zh/api-reference/image-series/models/nano-banana-2"
          }
        },
        {
          "label": {
            "en": "Nano Banana Pro",
            "zh": "Nano Banana Pro"
          },
          "href": {
            "en": "/en/api-reference/image-series/models/nano-banana-pro",
            "zh": "/zh/api-reference/image-series/models/nano-banana-pro"
          }
        }
      ],
      "body": {
        "en": "## Nano Banana\n\n**Nano Banana 2** and **Nano Banana Pro** are now available on OmniMux.\n\n### New models\n\n- **Model IDs:** `nano-banana-2`, `nano-banana-pro`\n- **Modality:** image generation\n\nSee each model page for request parameters.",
        "zh": "## Nano Banana\n\n**Nano Banana 2** 与 **Nano Banana Pro** 现已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `nano-banana-2`、`nano-banana-pro`\n- **模态：** 图像生成\n\n请求参数见各模型文档页。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-06-25-seedance-2-0",
      "published_at": "2026-06-25",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "Seedance 2.0 family — Standard, Mini, Fast",
        "zh": "Seedance 2.0 系列 — Standard / Mini / Fast"
      },
      "summary": {
        "en": "Seedance 2.0 video SKUs on OmniMux.",
        "zh": "Seedance 2.0 视频 SKU 已在 OmniMux 可调用。"
      },
      "models": [
        "seedance-2-0",
        "seedance-2-0-mini",
        "seedance-2-0-fast"
      ],
      "tags": [
        "video"
      ],
      "links": [
        {
          "label": {
            "en": "Video task poll",
            "zh": "视频任务查询"
          },
          "href": {
            "en": "/en/api-reference/tasks/video-task",
            "zh": "/zh/api-reference/tasks/video-task"
          }
        },
        {
          "label": {
            "en": "Pricing",
            "zh": "定价"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        }
      ],
      "body": {
        "en": "## Seedance 2.0\n\nThe **Seedance 2.0** family is available on OmniMux for async video generation.\n\n### New models\n\n| Model ID | Role |\n| --- | --- |\n| `seedance-2-0` | Standard |\n| `seedance-2-0-mini` | Lightweight / cost-oriented |\n| `seedance-2-0-fast` | Speed-oriented |\n\n### API\n\n- **Create:** `POST /v1/video/generations`\n- **Poll:** `GET /v1/video/generations/{task_id}`",
        "zh": "## Seedance 2.0\n\n**Seedance 2.0** 系列已在 OmniMux 可调用（异步视频）。\n\n### 新模型\n\n| 模型 ID | 角色 |\n| --- | --- |\n| `seedance-2-0` | 标准档 |\n| `seedance-2-0-mini` | 轻量 / 成本向 |\n| `seedance-2-0-fast` | 速度向 |\n\n### 调用\n\n- **创建：** `POST /v1/video/generations`\n- **查询：** `GET /v1/video/generations/{task_id}`"
      },
      "schema_version": 1
    },
    {
      "id": "2026-06-17-glm-5-2",
      "published_at": "2026-06-17",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "GLM-5.2",
        "zh": "GLM-5.2"
      },
      "summary": {
        "en": "GLM-5.2 language model on OmniMux.",
        "zh": "GLM-5.2 语言模型已在 OmniMux 可调用。"
      },
      "models": [
        "glm-5.2",
        "glm-5.1"
      ],
      "tags": [
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "GLM complete",
            "zh": "GLM 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/glm/complete",
            "zh": "/zh/api-reference/text-series/glm/complete"
          }
        }
      ],
      "body": {
        "en": "## GLM-5.2\n\n**GLM-5.2** is now available on OmniMux (with related GLM IDs on pricing).\n\n### New models\n\n- **Model IDs:** `glm-5.2`, `glm-5.1`\n- **API:** Chat Completions at `https://api.omnimux.ai/v1`",
        "zh": "## GLM-5.2\n\n**GLM-5.2** 现已在 OmniMux 可调用（定价上另有同族 GLM ID）。\n\n### 新模型\n\n- **模型 ID：** `glm-5.2`、`glm-5.1`\n- **API：** Chat Completions：`https://api.omnimux.ai/v1`"
      },
      "schema_version": 1
    },
    {
      "id": "2026-06-16-kling",
      "published_at": "2026-06-16",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "Kling V3 and Kling V2.6",
        "zh": "Kling V3 与 Kling V2.6"
      },
      "summary": {
        "en": "Kling video models on OmniMux.",
        "zh": "Kling 视频模型已在 OmniMux 可调用。"
      },
      "models": [
        "kling-v3",
        "kling-v2-6"
      ],
      "tags": [
        "video"
      ],
      "links": [
        {
          "label": {
            "en": "Video task poll",
            "zh": "视频任务查询"
          },
          "href": {
            "en": "/en/api-reference/tasks/video-task",
            "zh": "/zh/api-reference/tasks/video-task"
          }
        },
        {
          "label": {
            "en": "Pricing",
            "zh": "定价"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        }
      ],
      "body": {
        "en": "## Kling\n\n**Kling** video models are available on OmniMux.\n\n### New models\n\n- **Model IDs:** `kling-v3`, `kling-v2-6`\n- **API:** async `POST` / `GET` `/v1/video/generations`\n\nPin IDs from live pricing; request parameters follow the video task contract.",
        "zh": "## Kling\n\n**Kling** 视频模型已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `kling-v3`、`kling-v2-6`\n- **API：** 异步 `POST` / `GET` `/v1/video/generations`\n\n精确 ID 以线上定价为准；请求形态遵循视频任务约定。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-06-15-midjourney",
      "published_at": "2026-06-15",
      "type": "model_launch",
      "modality": [
        "image"
      ],
      "title": {
        "en": "Midjourney",
        "zh": "Midjourney"
      },
      "summary": {
        "en": "Midjourney image path on OmniMux live pricing.",
        "zh": "Midjourney 图像路径已在 OmniMux 线上定价可调用。"
      },
      "models": [
        "midjourney"
      ],
      "tags": [
        "image"
      ],
      "links": [
        {
          "label": {
            "en": "Pricing",
            "zh": "定价"
          },
          "href": {
            "en": "/en/api-reference/account/pricing",
            "zh": "/zh/api-reference/account/pricing"
          }
        }
      ],
      "body": {
        "en": "## Midjourney\n\n**Midjourney** is available on OmniMux as model ID `midjourney`.\n\n### New model\n\n- **Model ID:** `midjourney`\n- **Modality:** image generation via the gateway catalog\n\nConfirm request shape and billing on console pricing before production use.",
        "zh": "## Midjourney\n\n**Midjourney** 已在 OmniMux 可调用，模型 ID 为 `midjourney`。\n\n### 新模型\n\n- **模型 ID：** `midjourney`\n- **模态：** 网关目录中的图像生成\n\n上线生产前请在控制台定价确认请求形态与计费。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-06-01-minimax-m3",
      "published_at": "2026-06-01",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "MiniMax M3",
        "zh": "MiniMax M3"
      },
      "summary": {
        "en": "MiniMax M3 language model on OmniMux.",
        "zh": "MiniMax M3 语言模型已在 OmniMux 可调用。"
      },
      "models": [
        "minimax-m3"
      ],
      "tags": [
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "MiniMax complete",
            "zh": "MiniMax 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/minimax/complete",
            "zh": "/zh/api-reference/text-series/minimax/complete"
          }
        }
      ],
      "body": {
        "en": "## MiniMax M3\n\n**MiniMax M3** is now available on OmniMux Chat Completions.\n\n### New model\n\n- **Model ID:** `minimax-m3`\n- **API:** `https://api.omnimux.ai/v1`",
        "zh": "## MiniMax M3\n\n**MiniMax M3** 现已通过 Chat Completions 在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `minimax-m3`\n- **API：** `https://api.omnimux.ai/v1`"
      },
      "schema_version": 1
    },
    {
      "id": "2026-05-20-omni-flash",
      "published_at": "2026-05-20",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "Omni Flash — duration video family",
        "zh": "Omni Flash — 时长视频族"
      },
      "summary": {
        "en": "Omni Flash and duration SKUs on OmniMux.",
        "zh": "Omni Flash 及按时长 SKU 已在 OmniMux 可调用。"
      },
      "models": [
        "gemini-omni-flash"
      ],
      "tags": [
        "video"
      ],
      "links": [
        {
          "label": {
            "en": "Omni Flash",
            "zh": "Omni Flash"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/gemini-omni-flash",
            "zh": "/zh/api-reference/video-series/models/gemini-omni-flash"
          }
        },
        {
          "label": {
            "en": "Video task poll",
            "zh": "视频任务查询"
          },
          "href": {
            "en": "/en/api-reference/tasks/video-task",
            "zh": "/zh/api-reference/tasks/video-task"
          }
        }
      ],
      "body": {
        "en": "## Omni Flash\n\n**Omni Flash** (`gemini-omni-flash`) is available on OmniMux for fast video generation.\n\n### Model\n\n- **Model ID:** `gemini-omni-flash`\n- **API:** `POST /v1/video/generations` + poll `GET /v1/video/generations/{task_id}`",
        "zh": "## Omni Flash\n\n**Omni Flash**（`gemini-omni-flash`）已在 OmniMux 可调用，用于快速视频生成。\n\n### 模型\n\n- **模型 ID：** `gemini-omni-flash`\n- **调用：** `POST /v1/video/generations` + 轮询 `GET /v1/video/generations/{task_id}`"
      },
      "schema_version": 1
    },
    {
      "id": "2026-05-20-veo-3-1",
      "published_at": "2026-05-20",
      "type": "model_launch",
      "modality": [
        "video"
      ],
      "title": {
        "en": "Veo 3.1",
        "zh": "Veo 3.1"
      },
      "summary": {
        "en": "Veo 3.1 video generation on OmniMux.",
        "zh": "Veo 3.1 视频生成已在 OmniMux 可调用。"
      },
      "models": [
        "veo-3.1"
      ],
      "tags": [
        "video"
      ],
      "links": [
        {
          "label": {
            "en": "Veo 3.1",
            "zh": "Veo 3.1"
          },
          "href": {
            "en": "/en/api-reference/video-series/models/veo-3.1",
            "zh": "/zh/api-reference/video-series/models/veo-3.1"
          }
        },
        {
          "label": {
            "en": "Video task poll",
            "zh": "视频任务查询"
          },
          "href": {
            "en": "/en/api-reference/tasks/video-task",
            "zh": "/zh/api-reference/tasks/video-task"
          }
        }
      ],
      "body": {
        "en": "## Veo 3.1\n\n**Veo 3.1** is now available on OmniMux.\n\n### New model\n\n- **Model ID:** `veo-3.1`\n- **API:** async video create + poll on `/v1/video/generations`",
        "zh": "## Veo 3.1\n\n**Veo 3.1** 现已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `veo-3.1`\n- **API：** `/v1/video/generations` 异步创建 + 查询"
      },
      "schema_version": 1
    },
    {
      "id": "2026-05-05-claude-opus-4-7-deepseek-v4",
      "published_at": "2026-05-05",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "Claude Opus 4.7 and DeepSeek V4",
        "zh": "Claude Opus 4.7 与 DeepSeek V4"
      },
      "summary": {
        "en": "Claude Opus 4.7 plus DeepSeek V4 Flash/Pro on OmniMux.",
        "zh": "Claude Opus 4.7 与 DeepSeek V4 Flash/Pro 已在 OmniMux 可调用。"
      },
      "models": [
        "claude-opus-4-7",
        "deepseek-v4-flash",
        "deepseek-v4-pro"
      ],
      "tags": [
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "Claude complete",
            "zh": "Claude 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/claude/complete",
            "zh": "/zh/api-reference/text-series/claude/complete"
          }
        },
        {
          "label": {
            "en": "DeepSeek complete",
            "zh": "DeepSeek 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/deepseek/complete",
            "zh": "/zh/api-reference/text-series/deepseek/complete"
          }
        }
      ],
      "body": {
        "en": "## Claude Opus 4.7 and DeepSeek V4\n\nNew language models are available on OmniMux Chat Completions.\n\n### New models\n\n| Model ID | Brand |\n| --- | --- |\n| `claude-opus-4-7` | Claude |\n| `deepseek-v4-flash` | DeepSeek |\n| `deepseek-v4-pro` | DeepSeek |\n\n### API\n\n- **Endpoint:** `https://api.omnimux.ai/v1` (Chat Completions)",
        "zh": "## Claude Opus 4.7 与 DeepSeek V4\n\n新的语言模型已通过 Chat Completions 在 OmniMux 可调用。\n\n### 新模型\n\n| 模型 ID | 品牌 |\n| --- | --- |\n| `claude-opus-4-7` | Claude |\n| `deepseek-v4-flash` | DeepSeek |\n| `deepseek-v4-pro` | DeepSeek |\n\n### 调用\n\n- **端点：** `https://api.omnimux.ai/v1`（Chat Completions）"
      },
      "schema_version": 1
    },
    {
      "id": "2026-04-28-gpt-image-2",
      "published_at": "2026-04-28",
      "type": "model_launch",
      "modality": [
        "image"
      ],
      "title": {
        "en": "GPT Image 2",
        "zh": "GPT Image 2"
      },
      "summary": {
        "en": "GPT Image 2 and HD tier on OmniMux.",
        "zh": "GPT Image 2 及 HD 档已在 OmniMux 可调用。"
      },
      "models": [
        "gpt-image-2",
        "gpt-image-2-hd"
      ],
      "tags": [
        "image"
      ],
      "links": [
        {
          "label": {
            "en": "GPT Image 2",
            "zh": "GPT Image 2"
          },
          "href": {
            "en": "/en/api-reference/image-series/models/gpt-image-2",
            "zh": "/zh/api-reference/image-series/models/gpt-image-2"
          }
        },
        {
          "label": {
            "en": "GPT Image 2 HD",
            "zh": "GPT Image 2 HD"
          },
          "href": {
            "en": "/en/api-reference/image-series/models/gpt-image-2-hd",
            "zh": "/zh/api-reference/image-series/models/gpt-image-2-hd"
          }
        }
      ],
      "body": {
        "en": "## GPT Image 2\n\n**GPT Image 2** is now available on OmniMux.\n\n### New models\n\n- **Model IDs:** `gpt-image2`, `gpt-image2-hd`\n- **Modality:** image generation\n\nSee model docs for size / quality parameters.",
        "zh": "## GPT Image 2\n\n**GPT Image 2** 现已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `gpt-image2`、`gpt-image2-hd`\n- **模态：** 图像生成\n\n尺寸 / 画质等参数见模型文档。"
      },
      "schema_version": 1
    },
    {
      "id": "2026-03-08-minimax-m2-5",
      "published_at": "2026-03-08",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "MiniMax M2.5",
        "zh": "MiniMax M2.5"
      },
      "summary": {
        "en": "MiniMax M2.5 language model on OmniMux.",
        "zh": "MiniMax M2.5 语言模型已在 OmniMux 可调用。"
      },
      "models": [
        "minimax-m2.5",
        "minimax-m2.7"
      ],
      "tags": [
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "MiniMax complete",
            "zh": "MiniMax 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/minimax/complete",
            "zh": "/zh/api-reference/text-series/minimax/complete"
          }
        }
      ],
      "body": {
        "en": "## MiniMax M2.5\n\n**MiniMax M2.5** (and related MiniMax text IDs) are available on OmniMux.\n\n### New models\n\n- **Model IDs:** `minimax-m2.5`, `minimax-m2.7`\n- **API:** Chat Completions",
        "zh": "## MiniMax M2.5\n\n**MiniMax M2.5**（及同族文本 ID）已在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `minimax-m2.5`、`minimax-m2.7`\n- **API：** Chat Completions"
      },
      "schema_version": 1
    },
    {
      "id": "2026-03-06-gpt-5-4",
      "published_at": "2026-03-06",
      "type": "model_launch",
      "modality": [
        "text"
      ],
      "title": {
        "en": "GPT-5.4 series",
        "zh": "GPT-5.4 系列"
      },
      "summary": {
        "en": "GPT-5.4 and GPT-5.4 Mini on OmniMux.",
        "zh": "GPT-5.4 与 GPT-5.4 Mini 已在 OmniMux 可调用。"
      },
      "models": [
        "gpt-5.4",
        "gpt-5.4-mini"
      ],
      "tags": [
        "text"
      ],
      "links": [
        {
          "label": {
            "en": "GPT complete",
            "zh": "GPT 完整参数"
          },
          "href": {
            "en": "/en/api-reference/text-series/gpt/complete",
            "zh": "/zh/api-reference/text-series/gpt/complete"
          }
        }
      ],
      "body": {
        "en": "## GPT-5.4\n\n**GPT-5.4** series models are available on OmniMux Chat Completions.\n\n### New models\n\n- **Model IDs:** `gpt-5.4`, `gpt-5.4-mini`\n- **API:** `https://api.omnimux.ai/v1`\n\nLater GPT-5.x IDs (for example `gpt-5.5`, `gpt-5.6-terra`) may also appear on live pricing — always pin the ID from the console.",
        "zh": "## GPT-5.4\n\n**GPT-5.4** 系列模型已通过 Chat Completions 在 OmniMux 可调用。\n\n### 新模型\n\n- **模型 ID：** `gpt-5.4`、`gpt-5.4-mini`\n- **API：** `https://api.omnimux.ai/v1`\n\n后续 GPT-5.x ID（如 `gpt-5.5`、`gpt-5.6-terra`）也可能出现在线上定价——请始终以控制台 ID 为准。"
      },
      "schema_version": 1
    }
  ]
}