> ## Documentation Index
> Fetch the complete documentation index at: https://composio-preview.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# MiniMax

> MiniMax provides multimodal AI APIs for text, video, speech, image, music, and file workflows.

![MiniMax logo](https://mintcdn.com/composio-preview/9YPuzjEU4UlMnwGq/images/toolkits/minimax.svg?fit=max\&auto=format\&n=9YPuzjEU4UlMnwGq\&q=85\&s=a5a5f5588b9219163d0b398126021555)

* **Category:** artificial intelligence
* **Auth:** API\_KEY
* **Composio-managed OAuth available?** N/A
* **Tools:** 35
* **Triggers:** 0
* **Slug:** `MINIMAX`
* **Version:** 20261101\_00
* **Instant:** Supported on the latest version. Support varies by tool.

## Tools

> [accordion] Clone Voice
>
> **Slug:** `MINIMAX_CLONE_VOICE`
>
> Create a reusable MiniMax cloned voice from an uploaded voice sample, optionally generating paid preview audio.
>
> #### Input Parameters
>
> **text** `string`
>
> Optional paid preview text. When supplied, model is also required.
>
> **model** `string ("speech-2.8-hd" | "speech-2.8-turbo" | "speech-2.6-hd" | "speech-2.6-turbo" | "speech-02-hd" | "speech-02-turbo" | "speech-01-hd" | "speech-01-turbo")`
>
> Speech model for preview audio. Required with text.
>
> **file\_id** `string` *required*
>
> Uploaded voice\_clone audio file ID. Audio must be MP3, M4A, or WAV, 10 seconds to 5 minutes, and no larger than 20 MB.
>
> **accuracy** `number`
>
> ASR similarity threshold from 0 to 1; zero or omission uses provider default 0.7.
>
> **voice\_id** `string` *required*
>
> Unique custom voice ID, 8-256 characters, starting with a letter and not ending with '-' or '\_'.
>
> **clone\_prompt** `object`
>
> Optional short prompt sample and exact transcript to improve clone similarity.
>
> **aigc\_watermark** `boolean`
>
> Whether to append an AIGC watermark tone to preview audio.
>
> **language\_boost** `string ("Chinese" | "Chinese,Yue" | "English" | "Arabic" | "Russian" | "Spanish" | "French" | "Portuguese" | "German" | "Turkish" | "Dutch" | "Ukrainian" | "Vietnamese" | "Indonesian" | "Japanese" | "Italian" | "Korean" | "Thai" | "Polish" | "Romanian" | "Greek" | "Czech" | "Finnish" | "Hindi" | "Bulgarian" | "Danish" | "Hebrew" | "Malay" | "Persian" | "Slovak" | "Swedish" | "Croatian" | "Filipino" | "Hungarian" | "Norwegian" | "Slovenian" | "Catalan" | "Nynorsk" | "Tamil" | "Afrikaans" | "auto")`
>
> Documented language or dialect boost, or auto.
>
> **text\_validation** `string`
>
> Expected transcript used for optional ASR similarity validation.
>
> **need\_noise\_reduction** `boolean`
>
> Whether to apply noise reduction to the cloning sample.
>
> **need\_volume\_normalization** `boolean`
>
> Whether to normalize cloning-sample volume.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Count Message Tokens
>
> **Slug:** `MINIMAX_COUNT_MESSAGE_TOKENS`
>
> Count input tokens for an Anthropic-compatible message payload without generating content.
>
> #### Input Parameters
>
> **model** `string` *required*
>
> Text model ID whose tokenizer and request overhead to use. Call MINIMAX\_LIST\_TEXT\_MODELS to discover currently available IDs.
>
> **tools** `array`
>
> Anthropic-compatible tool definitions whose prompt overhead is counted.
>
> **system** `string`
>
> System prompt text or system content blocks included in the count.
>
> **messages** `array` *required*
>
> Messages in Anthropic format, including text or content blocks to count.
>
> **thinking** `object`
>
> Thinking configuration included in the token estimate.
>
> **tool\_choice** `object`
>
> Tool-selection configuration included in the token estimate.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Create Async Speech
>
> **Slug:** `MINIMAX_CREATE_ASYNC_SPEECH`
>
> Start a long-form asynchronous MiniMax speech synthesis task from inline text or one uploaded text file.
>
> #### Input Parameters
>
> **text** `string`
>
> Inline text to synthesize, up to 50,000 characters. Supply exactly one of text or text\_file\_id.
>
> **model** `string ("speech-2.8-hd" | "speech-2.8-turbo" | "speech-2.6-hd" | "speech-2.6-turbo" | "speech-02-hd" | "speech-02-turbo" | "speech-01-hd" | "speech-01-turbo")` *required*
>
> Speech synthesis model.
>
> **text\_file\_id** `string`
>
> Uploaded txt or zip file ID containing up to 1,000,000 characters. Supply exactly one of text\_file\_id or text.
>
> **voice\_modify** `object`
>
> Optional voice effects; supported only for mp3, wav, and flac output.
>
> **audio\_setting** `object`
>
> Generated audio encoding settings.
>
> **voice\_setting** `object` *required*
>
> Voice and delivery settings.
>
> **language\_boost** `string ("Chinese" | "Chinese,Yue" | "English" | "Arabic" | "Russian" | "Spanish" | "French" | "Portuguese" | "German" | "Turkish" | "Dutch" | "Ukrainian" | "Vietnamese" | "Indonesian" | "Japanese" | "Italian" | "Korean" | "Thai" | "Polish" | "Romanian" | "Greek" | "Czech" | "Finnish" | "Hindi" | "Bulgarian" | "Danish" | "Hebrew" | "Malay" | "Persian" | "Slovak" | "Swedish" | "Croatian" | "Filipino" | "Hungarian" | "Norwegian" | "Slovenian" | "Catalan" | "Nynorsk" | "Tamil" | "Afrikaans" | "auto")`
>
> Documented language or dialect boost value, or auto for detection.
>
> **pronunciation\_dict** `object`
>
> Custom pronunciation rules.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Create Legacy Video
>
> **Slug:** `MINIMAX_CREATE_LEGACY_VIDEO`
>
> Start a MiniMax v1 Hailuo text, image, first-and-last-frame, or subject-reference video task with mode-specific validation.
>
> #### Input Parameters
>
> **mode** `string ("text_to_video" | "image_to_video" | "first_last_frame" | "subject_reference")` *required*
>
> Generation mode. It selects the required input and valid model family and is not sent to MiniMax.
>
> **model** `string ("MiniMax-Hailuo-2.3" | "MiniMax-Hailuo-2.3-Fast" | "MiniMax-Hailuo-02" | "T2V-01-Director" | "T2V-01" | "I2V-01-Director" | "I2V-01-live" | "I2V-01" | "S2V-01")` *required*
>
> Legacy video model supported by the selected mode.
>
> **prompt** `string`
>
> Video prompt. Required for text\_to\_video and optional for the other modes.
>
> **duration** `integer ("6" | "10")`
>
> Video duration in seconds; supported combinations use 6 or 10.
>
> **resolution** `string ("512P" | "720P" | "768P" | "1080P")`
>
> Video resolution supported by the selected model, mode, and duration.
>
> **callback\_url** `string`
>
> HTTPS callback URL for task updates and challenge verification.
>
> **last\_frame\_image** `string`
>
> Public URL or base64 image data URI for the last frame. Used only by first\_last\_frame.
>
> **prompt\_optimizer** `boolean`
>
> Whether MiniMax may optimize the prompt.
>
> **fast\_pretreatment** `boolean`
>
> Whether to reduce prompt optimization time; supported only by Hailuo 2.x models.
>
> **first\_frame\_image** `string`
>
> Public URL or base64 image data URI for the first frame. Required by image modes.
>
> **subject\_reference** `array`
>
> Character reference used only by subject\_reference mode.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Create OpenAI-Compatible Response
>
> **Slug:** `MINIMAX_CREATE_RESPONSE`
>
> Call MiniMax through its OpenAI Responses compatibility contract and return one non-streaming response object.
>
> #### Input Parameters
>
> **text** `object`
>
> Text output configuration.
>
> **input** `string` *required*
>
> Text input or full Responses-compatible conversation history.
>
> **model** `string` *required*
>
> Model ID for the Responses-compatible request.
>
> **tools** `array`
>
> Available function tools.
>
> **top\_p** `number`
>
> Nucleus sampling probability.
>
> **stream** `boolean`
>
> Streaming is disabled because this action returns one JSON response.
>
> **metadata** `object`
>
> String-valued request metadata.
>
> **reasoning** `object`
>
> Reasoning configuration.
>
> **temperature** `number`
>
> Sampling temperature in the interval (0, 1].
>
> **tool\_choice** `string ("none" | "auto")`
>
> Tool selection strategy.
>
> **instructions** `string`
>
> System instructions.
>
> **service\_tier** `string ("standard" | "priority")`
>
> Request admission tier.
>
> **prompt\_cache\_key** `string`
>
> Prompt cache routing identifier.
>
> **max\_output\_tokens** `integer`
>
> Maximum output tokens, including reasoning tokens.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Create Template Video
>
> **Slug:** `MINIMAX_CREATE_TEMPLATE_VIDEO`
>
> Start a MiniMax Video Agent task using an official template and its ordered text and media inputs.
>
> #### Input Parameters
>
> **template\_id** `string` *required*
>
> Official MiniMax Video Agent template ID.
>
> **text\_inputs** `array`
>
> Text slot values in the order required by the template.
>
> **callback\_url** `string`
>
> HTTPS callback URL for task updates and challenge verification.
>
> **media\_inputs** `array`
>
> Media slot values in the order required by the template.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Create Video
>
> **Slug:** `MINIMAX_CREATE_VIDEO`
>
> Start a MiniMax H3 video v2 generation task from text, optionally using first/last-frame images or reference media. Returns an asynchronous task ID for MINIMAX\_GET\_VIDEO\_TASK.
>
> #### Input Parameters
>
> **extra** `object`
>
> Advanced MiniMax-H3-Max generation options, currently prompt expansion only.
>
> **model** `string ("MiniMax-H3" | "MiniMax-H3-Max")` *required*
>
> H3-family video model ID. MiniMax-H3 supports 768P and 2K; MiniMax-H3-Max supports 480P and 768P.
>
> **ratio** `string ("adaptive" | "21:9" | "16:9" | "4:3" | "1:1" | "3:4" | "9:16")`
>
> Output aspect ratio. Text-only generation requires a concrete ratio; image-to-video uses adaptive; reference generation defaults to adaptive.
>
> **content** `array` *required*
>
> Ordered multimodal content blocks. Include a non-empty text prompt and optionally first/last-frame images or reference image, video, and audio media. Frame and reference modes cannot be mixed.
>
> **duration** `integer` *required*
>
> Output duration in seconds. MiniMax-H3 supports 4-15; MiniMax-H3-Max supports 5-15.
>
> **resolution** `string ("480P" | "768P" | "2K")` *required*
>
> Output resolution. Use 768P or 2K with MiniMax-H3, and 480P or 768P with MiniMax-H3-Max.
>
> **callback\_url** `string`
>
> HTTPS callback URL for task updates. It must answer MiniMax's challenge verification within three seconds.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Create Video Prompt
>
> **Slug:** `MINIMAX_CREATE_VIDEO_PROMPT`
>
> Start a MiniMax H3 Context-IR task that converts multimodal context into an enhanced video prompt.
>
> #### Input Parameters
>
> **model** `string` *required*
>
> Context interpretation model; currently MiniMax-H3.
>
> **ratio** `string ("adaptive" | "21:9" | "16:9" | "4:3" | "1:1" | "3:4" | "9:16")`
>
> Intended aspect ratio. Text-only input requires a concrete ratio.
>
> **content** `array` *required*
>
> Prompt plus optional frame or reference media used to produce an enhanced prompt.
>
> **duration** `integer` *required*
>
> Intended video duration in seconds, from 4 through 15.
>
> **callback\_url** `string`
>
> HTTPS callback URL for task updates and challenge verification.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Delete File
>
> **Slug:** `MINIMAX_DELETE_FILE`
>
> Permanently delete one MiniMax file by its exact ID and storage-purpose category. This operation cannot be undone.
>
> #### Input Parameters
>
> **file\_id** `string` *required*
>
> Exact MiniMax file ID to permanently delete. Numeric IDs are accepted and preserved as strings at the tool boundary.
>
> **purpose** `string ("voice_clone" | "prompt_audio" | "t2a_async" | "t2a_async_input" | "video_generation")` *required*
>
> Purpose category under which MiniMax stores the file.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Cancel or Delete Video Task
>
> **Slug:** `MINIMAX_DELETE_VIDEO_TASK`
>
> Cancel a queued MiniMax video v2 task, or permanently delete a completed or failed task record. The operation performed depends on the task's current state.
>
> #### Input Parameters
>
> **task\_id** `string` *required*
>
> Video v2 task ID to cancel if the task is queued, or permanently delete if the task has completed or failed.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Delete Voice
>
> **Slug:** `MINIMAX_DELETE_VOICE`
>
> Permanently delete an eligible MiniMax cloned or designed voice.
>
> #### Input Parameters
>
> **voice\_id** `string` *required*
>
> ID of the cloned or generated voice to permanently delete.
>
> **voice\_type** `string ("voice_cloning" | "voice_generation")` *required*
>
> Deletable voice category: cloned or generated voice.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Design Voice
>
> **Slug:** `MINIMAX_DESIGN_VOICE`
>
> Design a reusable MiniMax voice from a description and return its paid preview as a downloadable audio file.
>
> #### Input Parameters
>
> **prompt** `string` *required*
>
> Natural-language description of the voice to design.
>
> **voice\_id** `string`
>
> Optional custom voice ID; MiniMax generates one when omitted.
>
> **preview\_text** `string` *required*
>
> Text spoken in the paid preview audio, up to 500 characters.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Download File
>
> **Slug:** `MINIMAX_DOWNLOAD_FILE`
>
> Download the binary contents of one generated MiniMax file and return a managed file reference. Response headers determine the original name and MIME type when available; downloads larger than 100 MB are rejected.
>
> #### Input Parameters
>
> **file\_id** `string` *required*
>
> Exact ID of the generated MiniMax file to download. Numeric IDs are accepted and preserved as strings.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Estimate Responses Input Tokens
>
> **Slug:** `MINIMAX_ESTIMATE_RESPONSE_INPUT_TOKENS`
>
> Estimate input tokens using MiniMax's OpenAI Responses compatibility contract without generating model output.
>
> #### Input Parameters
>
> **text** `object`
>
> Text output configuration included in the estimate.
>
> **input** `string` *required*
>
> Text input or full Responses-compatible conversation history.
>
> **model** `string` *required*
>
> Model ID whose Responses-compatible token accounting to use.
>
> **tools** `array`
>
> Function tools included in the estimate.
>
> **reasoning** `object`
>
> Reasoning configuration included in the estimate.
>
> **tool\_choice** `string ("none" | "auto")`
>
> Tool selection strategy.
>
> **instructions** `string`
>
> System instructions included in the estimate.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Generate Anthropic-Compatible Message
>
> **Slug:** `MINIMAX_GENERATE_ANTHROPIC_MESSAGE`
>
> Call MiniMax through its Anthropic Messages compatibility contract and return one non-streaming message response.
>
> #### Input Parameters
>
> **model** `string ("MiniMax-M3.1-Flash-Preview" | "MiniMax-M3" | "MiniMax-M2.7" | "MiniMax-M2.7-highspeed" | "MiniMax-M2.5" | "MiniMax-M2.5-highspeed" | "MiniMax-M2.1" | "MiniMax-M2.1-highspeed" | "MiniMax-M2")` *required*
>
> Model ID for the Anthropic-compatible request.
>
> **tools** `array`
>
> Available Anthropic-format tools.
>
> **top\_p** `number`
>
> Nucleus sampling probability.
>
> **stream** `boolean`
>
> Streaming is disabled because this action returns one JSON response.
>
> **system** `string`
>
> System prompt.
>
> **messages** `array` *required*
>
> Conversation messages in Anthropic format.
>
> **metadata** `object`
>
> Request metadata.
>
> **thinking** `object`
>
> Thinking configuration.
>
> **max\_tokens** `integer`
>
> Maximum generated tokens.
>
> **temperature** `number`
>
> Sampling temperature from 0 to 2.
>
> **tool\_choice** `object`
>
> Tool selection configuration.
>
> **service\_tier** `string ("standard" | "priority")`
>
> Request admission tier.
>
> **output\_config** `object`
>
> Output and thinking-depth settings.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Generate Image
>
> **Slug:** `MINIMAX_GENERATE_IMAGE`
>
> Generate one or more images from a text prompt, optionally guided by a subject reference.
>
> **Instant:** Supported on the latest version.
>
> **Instant pricing:** \$0.350 per 100 successfully generated image-01 images.
>
> #### Input Parameters
>
> **seed** `integer`
>
> Random seed. Reusing a seed with the same parameters helps reproduce an image.
>
> **model** `string ("image-01" | "image-01-live")` *required*
>
> Image generation model ID.
>
> **width** `integer`
>
> Output width in pixels (512-2,048, divisible by 8). Must be supplied with height.
>
> **height** `integer`
>
> Output height in pixels (512-2,048, divisible by 8). Must be supplied with width.
>
> **prompt** `string` *required*
>
> Detailed description of the image to generate (1-1,500 characters).
>
> **image\_count** `integer`
>
> Number of images to generate (1-9).
>
> **aspect\_ratio** `string ("1:1" | "16:9" | "4:3" | "3:2" | "2:3" | "3:4" | "9:16" | "21:9")`
>
> Preset output aspect ratio. If supplied with width and height, this preset takes priority.
>
> **response\_format** `string ("url" | "base64")`
>
> Return temporary image URLs or base64-encoded image data. URLs expire after 24 hours.
>
> **prompt\_optimizer** `boolean`
>
> Whether MiniMax may automatically optimize the prompt before generation.
>
> **subject\_reference** `array`
>
> Character portrait references for image-guided generation.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Generate Native Chat Completion
>
> **Slug:** `MINIMAX_GENERATE_NATIVE_CHAT_COMPLETION`
>
> Call MiniMax's native chatcompletion\_v2 contract and return one non-streaming completion, including native sensitivity metadata.
>
> #### Input Parameters
>
> **model** `string ("MiniMax-M3.1-Flash-Preview" | "MiniMax-M3" | "MiniMax-M2.7" | "MiniMax-M2.7-highspeed" | "MiniMax-M2.5" | "MiniMax-M2.5-highspeed" | "MiniMax-M2.1" | "MiniMax-M2")` *required*
>
> Model ID accepted by the native chatcompletion\_v2 contract.
>
> **tools** `array`
>
> Available function tools.
>
> **top\_p** `number`
>
> Nucleus sampling probability in the interval (0, 1].
>
> **stream** `boolean`
>
> Streaming is disabled because this action returns one JSON response.
>
> **messages** `array` *required*
>
> Conversation messages in MiniMax's native chat format.
>
> **max\_tokens** `integer`
>
> Deprecated maximum generated tokens.
>
> **temperature** `number`
>
> Sampling temperature in the interval (0, 1].
>
> **tool\_choice** `string ("none" | "auto")`
>
> Tool selection strategy.
>
> **stream\_options** `object`
>
> Streaming options retained for contract compatibility; stream is forced off.
>
> **response\_format** `object`
>
> Structured output format for supported models.
>
> **mask\_sensitive\_info** `boolean`
>
> Whether to mask sensitive information in output.
>
> **max\_completion\_tokens** `integer`
>
> Maximum generated tokens.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Generate Text
>
> **Slug:** `MINIMAX_GENERATE_TEXT`
>
> Generate one non-streaming OpenAI-compatible chat completion, including optional reasoning and tool calls.
>
> #### Input Parameters
>
> **model** `string` *required*
>
> Text model ID; use MINIMAX\_LIST\_TEXT\_MODELS to discover currently available IDs.
>
> **tools** `array`
>
> Function tools the model may call, using OpenAI-compatible tool definitions.
>
> **top\_p** `number`
>
> Nucleus-sampling probability mass from 0 through 1.
>
> **messages** `array` *required*
>
> Conversation messages in OpenAI chat format, each with a role and text or multimodal content.
>
> **thinking** `object`
>
> Thinking configuration for models that support selecting a mode.
>
> **temperature** `number`
>
> Sampling temperature from 0 through 2; higher values increase randomness.
>
> **reasoning\_effort** `string ("low" | "medium" | "high" | "xhigh" | "max")`
>
> Reasoning effort for supported models: low, medium, high, xhigh, or max.
>
> **max\_completion\_tokens** `integer`
>
> Maximum completion tokens; allow enough room for reasoning because a small value can end before visible output.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Account Balance
>
> **Slug:** `MINIMAX_GET_ACCOUNT_BALANCE`
>
> Get the connected global pay-as-you-go MiniMax account's available, cash, voucher, credit, owed, and balance-alert amounts.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Anthropic-Compatible Model
>
> **Slug:** `MINIMAX_GET_ANTHROPIC_MODEL`
>
> Retrieve one MiniMax text model through the Anthropic Models compatibility contract.
>
> #### Input Parameters
>
> **model\_id** `string` *required*
>
> Anthropic-compatible model identifier.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Async Speech Task
>
> **Slug:** `MINIMAX_GET_ASYNC_SPEECH_TASK`
>
> Get the status and generated file reference for a MiniMax asynchronous speech synthesis task.
>
> #### Input Parameters
>
> **task\_id** `string` *required*
>
> Async speech task ID returned by CREATE\_ASYNC\_SPEECH, as a decimal string.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get File
>
> **Slug:** `MINIMAX_GET_FILE`
>
> Retrieve metadata for one MiniMax file by its exact ID, including a temporary download URL when MiniMax provides one for a generated output.
>
> #### Input Parameters
>
> **file\_id** `string` *required*
>
> Exact MiniMax file ID returned by upload, video generation, or async speech synthesis. Numeric IDs are accepted and preserved as strings.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Legacy Video Task
>
> **Slug:** `MINIMAX_GET_LEGACY_VIDEO_TASK`
>
> Get the status and output file metadata of a MiniMax v1 legacy video task.
>
> #### Input Parameters
>
> **task\_id** `string` *required*
>
> Legacy task ID returned by MINIMAX\_CREATE\_LEGACY\_VIDEO.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Template Video Task
>
> **Slug:** `MINIMAX_GET_TEMPLATE_VIDEO_TASK`
>
> Get the status and temporary result URL of a MiniMax Video Agent template task.
>
> #### Input Parameters
>
> **task\_id** `string` *required*
>
> Template task ID returned by MINIMAX\_CREATE\_TEMPLATE\_VIDEO.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get OpenAI-Compatible Text Model
>
> **Slug:** `MINIMAX_GET_TEXT_MODEL`
>
> Retrieve one MiniMax text model through the OpenAI Models compatibility contract.
>
> #### Input Parameters
>
> **model\_id** `string` *required*
>
> OpenAI-compatible model identifier.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Get Video Task
>
> **Slug:** `MINIMAX_GET_VIDEO_TASK`
>
> Get the current state and output of one MiniMax video v2 task created within the last seven days. Failed tasks are returned as task records with error details.
>
> #### Input Parameters
>
> **task\_id** `string` *required*
>
> Video v2 task ID returned by MINIMAX\_CREATE\_VIDEO or MINIMAX\_LIST\_VIDEO\_TASKS.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] List Anthropic-Compatible Models
>
> **Slug:** `MINIMAX_LIST_ANTHROPIC_MODELS`
>
> List one page of MiniMax text models through the Anthropic Models compatibility contract, using its after\_id or before\_id cursors.
>
> #### Input Parameters
>
> **limit** `integer`
>
> Number of models to return on this page.
>
> **after\_id** `string`
>
> Return models after this model ID.
>
> **before\_id** `string`
>
> Return models before this model ID.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] List Files
>
> **Slug:** `MINIMAX_LIST_FILES`
>
> List MiniMax files for one upload-purpose category. MiniMax documents no pagination for this endpoint.
>
> #### Input Parameters
>
> **purpose** `string ("voice_clone" | "prompt_audio" | "t2a_async_input" | "video_generation_input")` *required*
>
> Purpose category whose uploaded files should be returned.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] List Text Models
>
> **Slug:** `MINIMAX_LIST_TEXT_MODELS`
>
> List text model IDs currently available to the connected MiniMax API key.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] List Video Tasks
>
> **Slug:** `MINIMAX_LIST_VIDEO_TASKS`
>
> List one page of MiniMax video v2 tasks from the last seven days, optionally filtered by status, task ID, model, or task type.
>
> #### Input Parameters
>
> **model** `string`
>
> Return only tasks created with this model ID.
>
> **cursor** `string`
>
> Opaque continuation cursor returned by the previous call. Omit for the first page and pass it back unchanged.
>
> **status** `string ("queued" | "running" | "succeeded" | "failed" | "cancelled")`
>
> Return only tasks in this status.
>
> **task\_ids** `array`
>
> Return only tasks with these task IDs.
>
> **page\_size** `integer`
>
> Maximum tasks to return in this page.
>
> **task\_type** `string ("generation" | "h3_context_ir" | "regeneration")`
>
> Return only tasks of this type.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] List Voices
>
> **Slug:** `MINIMAX_LIST_VOICES`
>
> List system, cloned, generated, or all voices that can be used for speech synthesis.
>
> #### Input Parameters
>
> **voice\_type** `string ("system" | "voice_cloning" | "voice_generation" | "all")` *required*
>
> Voice category to return: system for built-in voices, voice\_cloning for cloned voices, voice\_generation for designed voices, or all to return every category in one list.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Regenerate Video
>
> **Slug:** `MINIMAX_REGENERATE_VIDEO`
>
> Regenerate an eligible MiniMax-H3 768P source video as a 2K video, using either its task ID or the exact original inputs and source video.
>
> #### Input Parameters
>
> **model** `string` *required*
>
> Regeneration model; currently MiniMax-H3.
>
> **content** `array`
>
> Final source prompt and original generation inputs, plus exactly one video\_url item whose role is base\_video. Mutually exclusive with source\_task\_id.
>
> **resolution** `string` *required*
>
> Target regeneration resolution; currently 2K.
>
> **callback\_url** `string`
>
> HTTPS callback URL for task updates and challenge verification.
>
> **aigc\_watermark** `boolean`
>
> Whether to add an AIGC label watermark.
>
> **source\_task\_id** `string`
>
> Succeeded, owned v2 generation task ID from the last seven days. Mutually exclusive with content.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Synthesize Speech
>
> **Slug:** `MINIMAX_SYNTHESIZE_SPEECH`
>
> Synchronously synthesize up to 9,999 characters of text and return the decoded audio as a downloadable file.
>
> **Instant:** Supported on the latest version.
>
> **Instant pricing:** \$0.006 per 100 billable characters (speech-2.8-turbo); \$0.010 for speech-2.8-hd.
>
> #### Input Parameters
>
> **text** `string` *required*
>
> Text to synthesize (1-9,999 characters).
>
> **model** `string ("speech-2.8-hd" | "speech-2.8-turbo" | "speech-2.6-hd" | "speech-2.6-turbo" | "speech-02-hd" | "speech-02-turbo" | "speech-01-hd" | "speech-01-turbo")` *required*
>
> Speech synthesis model to use.
>
> **voice\_modify** `object`
>
> Optional voice effects supported for non-streaming mp3, wav, and flac output.
>
> **audio\_setting** `object`
>
> Generated audio encoding settings.
>
> **voice\_setting** `object` *required*
>
> Voice identity and delivery settings.
>
> **language\_boost** `string ("Chinese" | "Chinese,Yue" | "English" | "Arabic" | "Russian" | "Spanish" | "French" | "Portuguese" | "German" | "Turkish" | "Dutch" | "Ukrainian" | "Vietnamese" | "Indonesian" | "Japanese" | "Italian" | "Korean" | "Thai" | "Polish" | "Romanian" | "Greek" | "Czech" | "Finnish" | "Hindi" | "Bulgarian" | "Danish" | "Hebrew" | "Malay" | "Persian" | "Slovak" | "Swedish" | "Croatian" | "Filipino" | "Hungarian" | "Norwegian" | "Slovenian" | "Catalan" | "Nynorsk" | "Tamil" | "Afrikaans" | "auto")`
>
> Documented language or dialect boost value, or auto for detection; when omitted, MiniMax applies no language boost.
>
> **pronunciation\_dict** `object`
>
> Custom pronunciation rules applied simultaneously during synthesis.
>
> **text\_normalization\_mode** `string ("basic" | "quality")`
>
> Text normalization mode. Quality mode adds latency and is supported only by selected models.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Transcribe Audio
>
> **Slug:** `MINIMAX_TRANSCRIBE_AUDIO`
>
> Transcribe an uploaded audio file with optional language hints, timestamps, speaker diarization, subtitle output, or assembled streaming output.
>
> #### Input Parameters
>
> **file** `object` *required*
>
> Audio file to transcribe. Supported containers are WAV, AIFF, FLAC, ALAC/M4A, MP3, AAC, Opus, and OGG; maximum duration is 500 seconds and maximum size is 50 MB.
>
> **model** `string`
>
> MiniMax speech recognition model.
>
> **stream** `boolean`
>
> Stream incremental transcription events. Streaming supports only response\_format=json; the action assembles the deltas into final text.
>
> **language** `string ("" | "zh" | "yue" | "en" | "ja" | "ko" | "th" | "vi" | "id" | "ms" | "fil" | "ar" | "tr" | "fr" | "de" | "es" | "it" | "pt" | "pl" | "ru" | "uk")`
>
> Optional BCP-47 primary-language hint, such as en, zh, yue, ja, or es. Omit it for mixed-language recognition.
>
> **response\_format** `string ("json" | "verbose_json" | "srt" | "vtt")`
>
> Result format. json returns text and duration; verbose\_json also returns speaker/timestamp segments; srt and vtt return a downloadable subtitle file.
>
> **timestamp\_level** `string ("" | "sentence" | "word")`
>
> Timestamp granularity for verbose\_json, srt, or vtt. Empty or sentence uses sentence/segment timestamps; word uses character-level timestamps for Chinese and word-level timestamps for English. Ignored for json.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not

> [accordion] Upload File
>
> **Slug:** `MINIMAX_UPLOAD_FILE`
>
> Upload one file to MiniMax as multipart/form-data for voice cloning, async speech, video understanding, or video generation.
>
> #### Input Parameters
>
> **file** `object` *required*
>
> File to upload. Its original name and MIME type are preserved in the multipart file part; MiniMax validates purpose-specific format and size limits.
>
> **purpose** `string ("voice_clone" | "prompt_audio" | "t2a_async_input" | "video_understanding" | "video_generation_input")` *required*
>
> MiniMax ingestion purpose. voice\_clone and prompt\_audio accept MP3, M4A, or WAV; t2a\_async\_input accepts TXT or ZIP long-text input; video\_understanding accepts MP4, AVI, MOV, or MKV; video\_generation\_input accepts JPG/JPEG, PNG, WEBP, HEIC, or HEIF images up to 30 MB, MP4 or MOV video up to 50 MB, or WAV or MP3 audio up to 15 MB.
>
> #### Output
>
> **data** `string` *required*
>
> Data from the action execution
>
> **error** `string`
>
> Error if any occurred during the execution of the action
>
> **successful** `boolean` *required*
>
> Whether or not the action execution was successful or not
