MiniMax
内置当智能体需要使用 MiniMax 或 Hailuo 多模态 AI 时，可选择此集成：M3 和 M2.7 文本、聊天、响应和 token 估算；speech-2.8 或 speech-2.6 短文本及异步长文本转语音；语音克隆、设计和管理；Hailuo 2.3 文本、图像或参考内容转视频，以及带状态轮询的 Video Agent 模板任务；image-01 图像生成；music-3.0 歌曲、歌词和封面预处理；或模型和文件的列出、上传、检索和删除。若需要 MiniMax 专属模型，或需要一家提供商覆盖文本、音频、图像、视频和音乐，请选择此集成；当用户要求其他提供商时，请使用其他集成。需要 MiniMax API 密钥。进行付费生成、上传私有媒体、克隆或设计语音，或删除文件或语音前，须先询问。平台凭据仅限于经审核的、有成本上限的无状态操作，以及安全的公开元数据或状态读取。
让 AI 智能体使用 MiniMax
Aident Loadout 让 Claude Code、Codex、Cursor 和其他受支持的智能体调用 MiniMax。可选择 30 个受支持的操作。 可在 Audit 中审核每次执行。
MiniMax 使用 Aident 托管访问，无需添加自己的服务提供商凭证即可运行受支持的操作。
分类
Generative Ai价格与支付
- 价格
- 5 个付费操作的价格为 $0.01 / call + variable。
- 统一支付
- 暂不支持。需要时请连接服务提供商账号。
包含的操作
- Upload FileUpload private audio, text, or video inputs and list, inspect, download, or delete MiniMax account files used by speech, voice, video, and understanding workflows.
- Retrieve FileUpload private audio, text, or video inputs and list, inspect, download, or delete MiniMax account files used by speech, voice, video, and understanding workflows.
- List FilesUpload private audio, text, or video inputs and list, inspect, download, or delete MiniMax account files used by speech, voice, video, and understanding workflows.
- Retrieve File ContentUpload private audio, text, or video inputs and list, inspect, download, or delete MiniMax account files used by speech, voice, video, and understanding workflows.
- Delete FileUpload private audio, text, or video inputs and list, inspect, download, or delete MiniMax account files used by speech, voice, video, and understanding workflows.
- Image GenerationGenerate one or more image-01 images from text or a subject reference, with aspect-ratio, size, seed, and URL or base64 output controls.Variable pricing
- List ModelsList or retrieve MiniMax model metadata through the OpenAI-compatible and Anthropic-compatible model catalogs.
- Retrieve ModelList or retrieve MiniMax model metadata through the OpenAI-compatible and Anthropic-compatible model catalogs.
- Anthropic List ModelsList or retrieve MiniMax model metadata through the OpenAI-compatible and Anthropic-compatible model catalogs.
- Anthropic Retrieve ModelList or retrieve MiniMax model metadata through the OpenAI-compatible and Anthropic-compatible model catalogs.
- Generate MusicGenerate MiniMax Music 3.0 songs, create or edit lyrics, and preprocess private reference audio for cover workflows.Variable pricing
- Generate LyricsGenerate MiniMax Music 3.0 songs, create or edit lyrics, and preprocess private reference audio for cover workflows.$0.01 / call
- Cover PreprocessGenerate MiniMax Music 3.0 songs, create or edit lyrics, and preprocess private reference audio for cover workflows.
- T2a V2Create short synchronous or long asynchronous speech with MiniMax Speech 2.8, 2.6, and 02 voices, then poll long-form tasks for output file IDs.Variable pricing
- T2a Async V2Create short synchronous or long asynchronous speech with MiniMax Speech 2.8, 2.6, and 02 voices, then poll long-form tasks for output file IDs.
- T2a Async V2queryCreate short synchronous or long asynchronous speech with MiniMax Speech 2.8, 2.6, and 02 voices, then poll long-form tasks for output file IDs.
- Chat Completion V2Generate text with MiniMax M3 and M2-series models using native, OpenAI-compatible, Anthropic-compatible, or Responses APIs; estimate or count tokens before generation.
- Create ResponseGenerate text with MiniMax M3 and M2-series models using native, OpenAI-compatible, Anthropic-compatible, or Responses APIs; estimate or count tokens before generation.
- Estimate Input TokensGenerate text with MiniMax M3 and M2-series models using native, OpenAI-compatible, Anthropic-compatible, or Responses APIs; estimate or count tokens before generation.
- Chat Completion Open AiGenerate text with MiniMax M3 and M2-series models using native, OpenAI-compatible, Anthropic-compatible, or Responses APIs; estimate or count tokens before generation.
- Chat Completion AnthropicGenerate text with MiniMax M3 and M2-series models using native, OpenAI-compatible, Anthropic-compatible, or Responses APIs; estimate or count tokens before generation.
- Count Tokens AnthropicGenerate text with MiniMax M3 and M2-series models using native, OpenAI-compatible, Anthropic-compatible, or Responses APIs; estimate or count tokens before generation.
- Video Template GenerationCreate and poll MiniMax Video Agent template jobs that combine template-specific media and text inputs into generated videos.
- Query Video Template Generation TaskCreate and poll MiniMax Video Agent template jobs that combine template-specific media and text inputs into generated videos.
- Video GenerationCreate Hailuo video tasks from text, first or last frames, images, or subject references, then poll task status for the generated file ID.Variable pricing
- Query Video Generation TaskCreate Hailuo video tasks from text, first or last frames, images, or subject references, then poll task status for the generated file ID.
- Voice CloneClone a voice from uploaded audio, design a voice from a description, list available voices, or delete custom MiniMax voices.
- Voice DesignClone a voice from uploaded audio, design a voice from a description, list available voices, or delete custom MiniMax voices.
- Get VoiceClone a voice from uploaded audio, design a voice from a description, list available voices, or delete custom MiniMax voices.
- Delete VoiceClone a voice from uploaded audio, design a voice from a description, list available voices, or delete custom MiniMax voices.