MiniMax

https://platform.minimax.io/docs
Admin
MiniMax is an AI platform offering a suite of advanced models for polyglot programming mastery,...
Tokens:281,069
Snippets:1,926
Trust Score:9.8
Update:1 day ago
Show doc for...
Tokens:
Raw
### Install OpenCode

Source: https://platform.minimax.io/docs/m-plan/opencode

Manual setup step 1: install OpenCode via the official install script or npm. Use either command.

```bash
curl -fsSL https://opencode.ai/install | bash
# or: npm i -g opencode-ai
```

--------------------------------

### OpenAI-Compatible Node.js Example

Source: https://platform.minimax.io/docs/guides/text-generation

Node.js example using the OpenAI SDK to call MiniMax. Requires installing the OpenAI SDK first: npm install openai.

```javascript
// Please install the OpenAI SDK first: `npm install openai`
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.minimax.io/v1",
  apiKey: "<MINIMAX_API_KEY>",
});

const response = await client.chat.completions.create({
  model: "MiniMax-M3.1-Flash",
  reasoning_effort: "max",
  messages: [
    { role: "user", content: "Hi, how are you?" },
  ],
});

console.log(response.choices[0].message.content);
```

--------------------------------

### OpenAI-Compatible Python Example

Source: https://platform.minimax.io/docs/guides/text-generation

Python example using the OpenAI SDK to call MiniMax. Requires installing the OpenAI SDK first: pip install openai.

```python
# Please install the OpenAI SDK first: `pip install openai`
from openai import OpenAI

client = OpenAI(
    base_url="https://api.minimax.io/v1",
    api_key="<MINIMAX_API_KEY>",
)

response = client.chat.completions.create(
    model="MiniMax-M3.1-Flash",
    reasoning_effort="max",
    messages=[
        {"role": "user", "content": "Hi, how are you?"},
    ],
)

print(response.choices[0].message.content)
```

--------------------------------

### Anthropic-Compatible Node.js Example

Source: https://platform.minimax.io/docs/guides/text-generation

Node.js example using the Anthropic SDK to call MiniMax. Requires installing the Anthropic SDK first: npm install @anthropic-ai/sdk.

```javascript
// Please install the Anthropic SDK first: `npm install @anthropic-ai/sdk`
import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  baseURL: "https://api.minimax.io/anthropic",
  apiKey: "<MINIMAX_API_KEY>",
});

const message = await client.messages.create({
  model: "MiniMax-M3.1-Flash",
  output_config: { effort: "max" },
  max_tokens: 4096,
  messages: [
    { role: "user", content: "Hi, how are you?" },
  ],
});

for (const block of message.content) {
  if (block.type === "thinking") {
    console.log(`Thinking:\n${block.thinking}\n`);
  } else if (block.type === "text") {
    console.log(`Text:\n${block.text}\n`);
  }
}
```

--------------------------------

### Anthropic-Compatible Python Example

Source: https://platform.minimax.io/docs/guides/text-generation

Python example using the Anthropic SDK to call MiniMax. Requires installing the Anthropic SDK first: pip install anthropic.

```python
# Please install the Anthropic SDK first: `pip install anthropic`
import anthropic

client = anthropic.Anthropic(
    base_url="https://api.minimax.io/anthropic",
    api_key="<MINIMAX_API_KEY>",
)

message = client.messages.create(
    model="MiniMax-M3.1-Flash",
    output_config={"effort": "max"},
    max_tokens=4096,
    messages=[
        {"role": "user", "content": "Hi, how are you?"}
    ],
)

for block in message.content:
    if block.type == "thinking":
        print(f"Thinking:\n{block.thinking}\n")
    elif block.type == "text":
        print(f"Text:\n{block.text}\n")
```

--------------------------------

### Install and serve Open WebUI via pip

Source: https://platform.minimax.io/docs/m-plan/other-tools

Requires Python 3.11 to avoid compatibility issues. Run these commands to install and start the Open WebUI server.

```bash
pip install open-webui
open-webui serve
```

--------------------------------

### Start MiniMax Code CLI in project

Source: https://platform.minimax.io/docs/m-plan/minimax-code

Run from your project directory. Use `/status` to check account, `/provider` to choose a model, and describe your development task.

```bash
mcode
```

--------------------------------

### Go example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses net/http to POST to the Get Voice endpoint and prints the response body. Replace <token> and <content-type> with your credentials.

```go
package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/get_voice"

	payload := strings.NewReader("{\n  \"voice_type\": \"all\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "<content-type>")
	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}
```

--------------------------------

### Go example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses net/http to POST to the Get Voice endpoint with voice_type 'all'. Prints the response body.

```go
package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/get_voice"

	payload := strings.NewReader("{\n  \"voice_type\": \"all\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "<content-type>")
	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}

```

--------------------------------

### Install MiniMax Code CLI on macOS, Linux, WSL

Source: https://platform.minimax.io/docs/m-plan/minimax-code

Official installer for macOS, Linux, and WSL. Run in terminal, then reopen terminal and verify with `mcode --version`.

```bash
curl -fsSL https://filecdn.minimax.chat/public/install.sh | bash
```

--------------------------------

### Agent Setup Command: mmx agent setup

Source: https://platform.minimax.io/docs/m-plan/minimax-cli

Use this command to configure MiniMax in AI coding tools.

```bash
mmx agent setup
```

--------------------------------

### Install MiniMax Code CLI on Windows (PowerShell)

Source: https://platform.minimax.io/docs/m-plan/minimax-code

Official installer for Windows PowerShell. Run in PowerShell, then reopen terminal and verify with `mcode --version`.

```powershell
irm https://filecdn.minimax.chat/public/install.ps1 | iex
```

--------------------------------

### Install Pi (macOS/Linux)

Source: https://platform.minimax.io/docs/m-plan/minimax-cli

Official installer command used by the wizard on macOS and Linux for Pi.

```bash
npm install -g --ignore-scripts --engine-strict @earendil-works/pi-coding-agent
```

--------------------------------

### Sign in to MiniMax CLI

Source: https://platform.minimax.io/docs/m-plan/minimax-cli

Sign in using your Subscription Key. Replace `sk-xxxxx` with your actual key. MiniMax CLI detects the service region from your key. After signing in, run `mmx quota` to confirm setup; if API calls return 401, see Troubleshooting.

```bash
mmx auth login --api-key sk-xxxxx
```

--------------------------------

### Verify Node.js and npm Installation

Source: https://platform.minimax.io/docs/guides/mcp-guide

Run these commands to confirm Node.js and npm are installed; they display the installed versions.

```bash
node -v 
npm -v
```

--------------------------------

### List Files with Go

Source: https://platform.minimax.io/docs/api-reference/file-management-list

Uses net/http to GET the List Files endpoint and prints the response body.

```go
package main

import (
	"fmt"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/files/list"

	req, _ := http.NewRequest("GET", url, nil)

	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}
```

--------------------------------

### Install OpenCode (macOS/Linux)

Source: https://platform.minimax.io/docs/m-plan/minimax-cli

Official installer command used by the wizard on macOS and Linux for OpenCode.

```bash
npm install -g opencode-ai
```

--------------------------------

### task_start

Source: https://platform.minimax.io/docs/api-reference/speech-t2a-websocket

Sends the task_start event to begin a speech synthesis task. The task is considered started when the server returns a task_started event, after which task_continue or task_finish events can be sent.

```APIDOC
## task_start

### Description
Sends the task_start event to begin a speech synthesis task. The task is considered started when the server returns a task_started event, after which task_continue or task_finish events can be sent.

### Event
`task_start`

### Parameters
- **model** (string) - Required - The model to use, e.g., "speech-2.8-turbo"
- **language_boost** (string) - Optional - Language boost, e.g., "Chinese"
- **voice_setting** (object) - Required - Voice configuration
  - **voice_id** (string) - Required - Voice identifier, e.g., "English_expressive_narrator"
  - **speed** (number) - Optional - Speech speed, default 1
  - **vol** (number) - Optional - Volume, default 1
  - **pitch** (number) - Optional - Pitch, default 0
- **pronunciation_dict** (object) - Optional - Pronunciation dictionary
  - **tone** (array) - Optional - Tone entries, e.g., ["Omg/Oh my god"]
- **audio_setting** (object) - Required - Audio output settings
  - **sample_rate** (number) - Required - Sample rate, e.g., 32000
  - **bitrate** (number) - Required - Bitrate, e.g., 128000
  - **format** (string) - Required - Audio format, e.g., "mp3"
  - **channel** (number) - Required - Number of channels, e.g., 1

### Request Example
```json
{
  "event": "task_start",
  "model": "speech-2.8-turbo",
  "language_boost": "Chinese",
  "voice_setting": {
    "voice_id": "English_expressive_narrator",
    "speed": 1,
    "vol": 1,
    "pitch": 0
  },
  "pronunciation_dict": {
    "tone": [
      "Omg/Oh my god"
    ]
  },
  "audio_setting": {
    "sample_rate": 32000,
    "bitrate": 128000,
    "format": "mp3",
    "channel": 1
  }
}
```

### Response
#### Success Response
- **event** (string) - "task_started"
- **session_id** (string) - Session identifier
- **trace_id** (string) - Trace identifier
- **base_resp** (object) - Base response
  - **status_code** (number) - Status code, 0 for success
  - **status_msg** (string) - Status message, "success" for success

#### Response Example
```json
{
  "session_id": "xxxx",
  "event": "task_started",
  "trace_id": "0303a2882bf18235ae7a809ae0f3cca7",
  "base_resp": {
    "status_code": 0,
    "status_msg": "success"
  }
}
```
```

--------------------------------

### Verify Node.js installation

Source: https://platform.minimax.io/docs/m-plan/minimax-cli

Run these commands after installing Node.js to confirm the installation. If both commands show a version number, the installation was successful.

```bash
node --version
npx --version
```

--------------------------------

### Python requests example to retrieve model

Source: https://platform.minimax.io/docs/api-reference/models/openai/retrieve-model

Uses the requests library to perform a GET request to the model endpoint. Prints the response text. Requires the requests library to be installed.

```python
import requests

url = "https://api.minimax.io/v1/models/{model_id}"

headers = {"Authorization": "Bearer <token>"}

response = requests.get(url, headers=headers)

print(response.text)
```

--------------------------------

### Initialize nanobot with onboard

Source: https://platform.minimax.io/docs/m-plan/other-tools

Run the nanobot onboard command to initialize the agent.

```bash
nanobot onboard
```

--------------------------------

### List Models with Go HTTP client

Source: https://platform.minimax.io/docs/api-reference/models/openai/list-models

Uses Go's net/http package to send a GET request to the models endpoint. Prints the response body.

```go
package main

import (
	"fmt"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/models"

	req, _ := http.NewRequest("GET", url, nil)

	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}
```

--------------------------------

### Install LangChain OpenAI adapter

Source: https://platform.minimax.io/docs/m-plan/other-tools

Install the `langchain-openai` package to use MiniMax via LangChain's OpenAI-compatible adapter.

```bash
pip install langchain-openai
```

--------------------------------

### Launch nanobot agent

Source: https://platform.minimax.io/docs/m-plan/other-tools

Start the nanobot agent after configuration.

```bash
nanobot agent
```

--------------------------------

### 200 response example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Example successful response with status code 200. Contains system_voice, voice_cloning, voice_generation arrays and base_resp with status_code 0 and status_msg 'success'.

```json
{
  "system_voice": [
    {
      "voice_id": "Chinese (Mandarin)_Reliable_Executive",
      "description": [
        "A steady and reliable male executive voice in standard Mandarin, conveying a trustworthy impression."
      ],
      "voice_name": "Steady Executive",
      "created_time": "1970-01-01"
    },
    {
      "voice_id": "Chinese (Mandarin)_News_Anchor",
      "description": [
        "A professional middle-aged female news anchor voice in standard Mandarin, with a broadcast-style tone."
      ],
      "voice_name": "News Anchor (Female)",
      "created_time": "1970-01-01"
    }
  ],
  "voice_cloning": [
    {
      "voice_id": "test12345",
      "description": [],
      "created_time": "2025-08-20"
    },
    {
      "voice_id": "test12346",
      "description": [],
      "created_time": "2025-08-21"
    }
  ],
  "voice_generation": [
    {
      "voice_id": "ttv-voice-2025082011321125-2uEN0X1S",
      "description": [],
      "created_time": "2025-08-20"
    },
    {
      "voice_id": "ttv-voice-2025082014225025-ZCQt0U0k",
      "description": [],
      "created_time": "2025-08-20"
    }
  ],
  "base_resp": {
    "status_code": 0,
    "status_msg": "success"
  }
}
```

--------------------------------

### List Files with Ruby Net::HTTP

Source: https://platform.minimax.io/docs/api-reference/file-management-list

Uses Net::HTTP to GET the List Files endpoint and prints the response body.

```ruby
require 'uri'
require 'net/http'

url = URI("https://api.minimax.io/v1/files/list")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Get.new(url)
request["Authorization"] = 'Bearer <token>'

response = http.request(request)
puts response.read_body
```

--------------------------------

### Start the SGLang server for MiniMax-M3

Source: https://platform.minimax.io/docs/guides/local-deploy-m3

Starts the SGLang server with the MiniMax-M3 model, using tensor parallelism across 8 GPUs and specific attention and MoE backends. The first start downloads approximately 444 GB of weights and may take a long time. The server listens on all interfaces inside the container, but is published only on the host loopback interface via -p 127.0.0.1:30000:30000.

```bash
docker run --rm --name minimax-m3 \
  --gpus all \
  --shm-size 32g \
  --ipc=host \
  -p 127.0.0.1:30000:30000 \
  -v ~/.cache/huggingface:/root/.cache/huggingface \
  lmsysorg/sglang@sha256:de63ac56df5d7b064451e21147eaab89634a02332d830ca8c01cb8c033b3a78f \
  sglang serve \
  --trust-remote-code \
  --model-path MiniMaxAI/MiniMax-M3-MXFP8 \
  --revision c5454eb03678d8710e54a4e0fc681b9f3b4a3dba \
  --reasoning-parser auto \
  --tool-call-parser auto \
  --tp 8 \
  --attention-backend fa4 \
  --mm-attention-backend flashinfer_cudnn \
  --moe-runner-backend deep_gemm \
  --chunked-prefill-size 8192 \
  --mem-fraction-static 0.65 \
  --host 0.0.0.0 \
  --port 30000
```

--------------------------------

### Start SGLang-Omni service on loopback

Source: https://platform.minimax.io/docs/guides/local-deploy-music-3

Launches the SGLang-Omni server for Music 3 on GPU 0, listening only on 127.0.0.1:8000. Keep this terminal running; the default configuration binds to the loopback interface only.

```bash
CUDA_VISIBLE_DEVICES=0 sgl-omni serve \
  --model-path "$MUSIC3_MODEL_DIR" \
  --model-name MiniMaxAI/MiniMax-Music3 \
  --host 127.0.0.1 \
  --port 8000
```

--------------------------------

### Java Unirest example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses Unirest to POST to the Get Voice endpoint with voice_type 'all'. Returns the response as a string.

```java
HttpResponse<String> response = Unirest.post("https://api.minimax.io/v1/get_voice")
  .header("Content-Type", "<content-type>")
  .header("Authorization", "Bearer <token>")
  .body("{\n  \"voice_type\": \"all\"\n}")
  .asString();

```

--------------------------------

### JavaScript fetch example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses fetch to POST to the Get Voice endpoint with voice_type 'all'. Logs the JSON response or error.

```javascript
const options = {
  method: 'POST',
  headers: {'Content-Type': '<content-type>', Authorization: 'Bearer <token>'},
  body: JSON.stringify({voice_type: 'all'})
};

fetch('https://api.minimax.io/v1/get_voice', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));

```

--------------------------------

### Launch OpenHands

Source: https://platform.minimax.io/docs/m-plan/other-tools

Start the OpenHands application after configuration.

```bash
openhands
```

--------------------------------

### Launch Pi after setup

Source: https://platform.minimax.io/docs/m-plan/pi

Launches Pi directly after the setup wizard completes.

```bash
pi
```

--------------------------------

### Python requests example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses the requests library to POST to the Get Voice endpoint with voice_type 'all'. Prints the response text.

```python
import requests

url = "https://api.minimax.io/v1/get_voice"

payload = { "voice_type": "all" }
headers = {
    "Content-Type": "<content-type>",
    "Authorization": "Bearer <token>"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)

```

--------------------------------

### Go example for music_cover_preprocess

Source: https://platform.minimax.io/docs/api-reference/music-cover-preprocess

Uses net/http to POST a JSON payload to the endpoint and prints the response body. Requires Go's standard library.

```go
package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/music_cover_preprocess"

	payload := strings.NewReader("{\n  \"model\": \"music-cover\",\n  \"audio_url\": \"https://example.com/song.mp3\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "<content-type>")
	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}
```

--------------------------------

### Install Grok CLI (macOS/Linux)

Source: https://platform.minimax.io/docs/m-plan/minimax-cli

Official installer command used by the wizard on macOS and Linux for Grok CLI.

```bash
curl -fsSL https://x.ai/cli/install.sh | bash
```

--------------------------------

### Ruby Net::HTTP example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses Net::HTTP to POST to the Get Voice endpoint with voice_type 'all'. Prints the response body.

```ruby
require 'uri'
require 'net/http'

url = URI("https://api.minimax.io/v1/get_voice")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Content-Type"] = '<content-type>'
request["Authorization"] = 'Bearer <token>'
request.body = "{\n  \"voice_type\": \"all\"\n}"

response = http.request(request)
puts response.read_body

```

--------------------------------

### PHP cURL example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses PHP cURL to POST to the Get Voice endpoint with voice_type 'all'. Outputs the response or a cURL error.

```php
<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.minimax.io/v1/get_voice",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'voice_type' => 'all'
  ]),
  CURLOPT_HTTPHEADER => [
    "Authorization: Bearer <token>",
    "Content-Type: <content-type>"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}

```

--------------------------------

### Sign in to MiniMax CLI

Source: https://platform.minimax.io/docs/m-plan/minimax-code

Run this command and complete sign-in in your browser. Required before using the CLI.

```bash
mcode login --region global
```

--------------------------------

### Java Unirest example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses Unirest to POST to the Get Voice endpoint and returns the response as a string. Replace <token> and <content-type> with your credentials.

```java
HttpResponse<String> response = Unirest.post("https://api.minimax.io/v1/get_voice")
  .header("Content-Type", "<content-type>")
  .header("Authorization", "Bearer <token>")
  .body("{\n  \"voice_type\": \"all\"\n}")
  .asString();
```

--------------------------------

### JavaScript fetch example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses fetch to POST to the Get Voice endpoint and logs the JSON response. Replace <token> and <content-type> with your credentials.

```javascript
const options = {
  method: 'POST',
  headers: {'Content-Type': '<content-type>', Authorization: 'Bearer <token>'},
  body: JSON.stringify({voice_type: 'all'})
};

fetch('https://api.minimax.io/v1/get_voice', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));
```

--------------------------------

### Python requests example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses the requests library to POST to the Get Voice endpoint. Prints the response text. Replace <token> and <content-type> with your credentials.

```python
import requests

url = "https://api.minimax.io/v1/get_voice"

payload = { "voice_type": "all" }
headers = {
    "Content-Type": "<content-type>",
    "Authorization": "Bearer <token>"
}

response = requests.post(url, json=payload, headers=headers)

print(response.text)
```

--------------------------------

### Install OpenHands with pipx

Source: https://platform.minimax.io/docs/m-plan/other-tools

Install the OpenHands CLI tool using pipx. This is the first step before configuring MiniMax.

```bash
pipx install openhands
```

--------------------------------

### Go example for music generation

Source: https://platform.minimax.io/docs/api-reference/music-generation

Uses net/http to POST the payload and prints the response body. Requires Go and valid credentials.

```go
package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/music_generation"

	payload := strings.NewReader("{\n  \"model\": \"music-3.0\",\n  \"prompt\": \"Indie folk, melancholic, introspective, longing, solitary walk, coffee shop\",\n  \"lyrics\": \"[verse]\\nStreetlights flicker, the night breeze sighs\\nShadows stretch as I walk alone\\nAn old coat wraps my silent sorrow\\nWandering, longing, where should I go\\n[chorus]\\nPushing the wooden door, the aroma spreads\\nIn a familiar corner, a stranger gazes\",\n  \"audio_setting\": {\n    \"sample_rate\": 44100,\n    \"bitrate\": 256000,\n    \"format\": \"mp3\"\n  }\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "<content-type>")
	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}
```

--------------------------------

### Ruby Net::HTTP example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses Net::HTTP to POST to the Get Voice endpoint and prints the response body. Replace <token> and <content-type> with your credentials.

```ruby
require 'uri'
require 'net/http'

url = URI("https://api.minimax.io/v1/get_voice")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Content-Type"] = '<content-type>'
request["Authorization"] = 'Bearer <token>'
request.body = "{\n  \"voice_type\": \"all\"\n}"

response = http.request(request)
puts response.read_body
```

--------------------------------

### Use MiniMax with LangChain ChatOpenAI

Source: https://platform.minimax.io/docs/m-plan/other-tools

Create a ChatOpenAI instance configured for MiniMax-M3.1-Flash. Replace `sk-cp-...` with your Subscription Key. The example invokes the model with 'Hello' and prints the response content.

```python
from langchain_openai import ChatOpenAI

llm = ChatOpenAI(
    model="MiniMax-M3.1-Flash",
    api_key="sk-cp-...",
    base_url="https://api.minimax.io/v1",
)
print(llm.invoke("Hello").content)
```

--------------------------------

### PHP cURL example for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Uses PHP cURL to POST to the Get Voice endpoint. Prints the response or a cURL error. Replace <token> and <content-type> with your credentials.

```php
<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.minimax.io/v1/get_voice",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "POST",
  CURLOPT_POSTFIELDS => json_encode([
    'voice_type' => 'all'
  ]),
  CURLOPT_HTTPHEADER => [
    "Authorization: Bearer <token>",
    "Content-Type: <content-type>"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}
```

--------------------------------

### Start FL2VA service

Source: https://platform.minimax.io/docs/guides/local-deploy-h3

Starts the SGLang server with the FL2VA model variant, using 8 GPUs with ulysses parallelism. Run from the SGLang repository with the virtual environment active. First startup loads the model and performs warmup, so readiness can take several minutes.

```bash
HF_HUB_OFFLINE=1 sglang serve \
  --model-path "$MINIMAX_H3_MODEL_DIR" \
  --model-variant fl2va \
  --num-gpus 8 \
  --tp-size 1 \
  --ulysses-degree 8 \
  --encoder-parallel auto \
  --performance-mode speed \
  --host 127.0.0.1 \
  --port 30010
```

--------------------------------

### First Request - Establish Cache (OpenAI SDK)

Source: https://platform.minimax.io/docs/api-reference/text-prompt-caching

Send the first request with the full content to create a cache entry. Uses reasoning_split=True to separate thinking content into reasoning_details field.

```python
from openai import OpenAI

client = OpenAI()

response1 = client.chat.completions.create(
    model="MiniMax-M3.1-Flash",
    messages=[
        {"role": "system", "content": "You are an AI assistant tasked with analyzing literary works. Your goal is to provide insightful commentary on themes, characters, and writing style.\n"},
        {"role": "user", "content": "<the entire contents of 'Pride and Prejudice'>"},
    ],
    # Set reasoning_split=True to separate thinking content into reasoning_details field
    extra_body={"reasoning_split": True},
)

print("First request result:")
print(f"Response: {response1.choices[0].message.content}")
print(f"Total Tokens: {response1.usage.total_tokens}")
print(f"Cached Tokens: {response1.usage.prompt_tokens_details.cached_tokens if hasattr(response1.usage, 'prompt_tokens_details') else 0}")
```

--------------------------------

### Task Started Event JSON response

Source: https://platform.minimax.io/docs/api-reference/speech-t2a-websocket

Received when the T2A task has started successfully. Only after this event can you send `task_continue` or `task_finish` events.

```json
{
  "session_id": "xxxx",
  "event": "task_started",
  "trace_id": "0303a2882bf18235ae7a809ae0f3cca7",
  "base_resp": {
    "status_code": 0,
    "status_msg": "success"
  }
}
```

--------------------------------

### Install nanobot via uv, pipx, or pip

Source: https://platform.minimax.io/docs/m-plan/other-tools

Install the nanobot CLI. The pip command may need --user or --break-system-packages on macOS.

```bash
uv tool install nanobot-ai
# or:  pipx install nanobot-ai
# or:  pip install nanobot-ai          # may need --user or --break-system-packages on macOS
```

--------------------------------

### cURL GET request to retrieve file

Source: https://platform.minimax.io/docs/api-reference/file-management-retrieve

Sends a GET request to the retrieve file endpoint with bearer token authorization. Replace <token> with your actual API token.

```bash
curl --request GET \
  --url https://api.minimax.io/v1/files/retrieve \
  --header 'Authorization: Bearer <token>'
```

--------------------------------

### Go example for video generation

Source: https://platform.minimax.io/docs/api-reference/video-generation-fl2v

Uses net/http to POST the video generation request with a JSON string body. Prints the response body.

```go
package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/video_generation"

	payload := strings.NewReader("{\n  \"prompt\": \"A little girl grow up.\",\n  \"first_frame_image\": \"https://filecdn.minimax.chat/public/fe9d04da-f60e-444d-a2e0-18ae743add33.jpeg\",\n  \"last_frame_image\": \"https://filecdn.minimax.chat/public/97b7cd08-764e-4b8b-a7bf-87a0bd898575.jpeg\",\n  \"model\": \"MiniMax-Hailuo-02\",\n  \"duration\": 6,\n  \"resolution\": \"1080P\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "<content-type>")
	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}
```

--------------------------------

### Successful response (200) for Get Voice

Source: https://platform.minimax.io/docs/api-reference/voice-management-get

Example 200 response listing system voices, voice cloning, voice generation, and base_resp with status_code 0 and status_msg 'success'.

```json
{
  "system_voice": [
    {
      "voice_id": "Chinese (Mandarin)_Reliable_Executive",
      "description": [
        "A steady and reliable male executive voice in standard Mandarin, conveying a trustworthy impression."
      ],
      "voice_name": "Steady Executive",
      "created_time": "1970-01-01"
    },
    {
      "voice_id": "Chinese (Mandarin)_News_Anchor",
      "description": [
        "A professional middle-aged female news anchor voice in standard Mandarin, with a broadcast-style tone."
      ],
      "voice_name": "News Anchor (Female)",
      "created_time": "1970-01-01"
    }
  ],
  "voice_cloning": [
    {
      "voice_id": "test12345",
      "description": [],
      "created_time": "2025-08-20"
    },
    {
      "voice_id": "test12346",
      "description": [],
      "created_time": "2025-08-21"
    }
  ],
  "voice_generation": [
    {
      "voice_id": "ttv-voice-2025082011321125-2uEN0X1S",
      "description": [],
      "created_time": "2025-08-20"
    },
    {
      "voice_id": "ttv-voice-2025082014225025-ZCQt0U0k",
      "description": [],
      "created_time": "2025-08-20"
    }
  ],
  "base_resp": {
    "status_code": 0,
    "status_msg": "success"
  }
}

```

--------------------------------

### Launch Oh My Pi

Source: https://platform.minimax.io/docs/m-plan/other-tools

Launch Oh My Pi directly after the setup wizard completes.

```bash
omp
```

--------------------------------

### Java Unirest GET to retrieve file

Source: https://platform.minimax.io/docs/api-reference/file-management-retrieve

Uses Unirest to send a GET request to the retrieve file endpoint and returns the response as a string. Replace <token> with your API token.

```java
HttpResponse<String> response = Unirest.get("https://api.minimax.io/v1/files/retrieve")
  .header("Authorization", "Bearer <token>")
  .asString();
```

--------------------------------

### Go example for lyrics generation

Source: https://platform.minimax.io/docs/api-reference/lyrics-generation

Creates an HTTP POST request with the lyrics payload and prints the response body. Uses net/http and strings.NewReader for the JSON body.

```go
package main

import (
	"fmt"
	"strings"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/lyrics_generation"

	payload := strings.NewReader("{\n  \"mode\": \"write_full_song\",\n  \"prompt\": \"A cheerful love song about a summer day at the beach\"\n}")

	req, _ := http.NewRequest("POST", url, payload)

	req.Header.Add("Content-Type", "<content-type>")
	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}
```

--------------------------------

### Go HTTP GET to retrieve file

Source: https://platform.minimax.io/docs/api-reference/file-management-retrieve

Uses net/http to send a GET request to the retrieve file endpoint and prints the response body. Replace <token> with your API token.

```go
package main

import (
	"fmt"
	"net/http"
	"io"
)

func main() {

	url := "https://api.minimax.io/v1/files/retrieve"

	req, _ := http.NewRequest("GET", url, nil)

	req.Header.Add("Authorization", "Bearer <token>")

	res, _ := http.DefaultClient.Do(req)

	defer res.Body.Close()
	body, _ := io.ReadAll(res.Body)

	fmt.Println(string(body))

}
```

--------------------------------

### JavaScript fetch example for music_cover_preprocess

Source: https://platform.minimax.io/docs/api-reference/music-cover-preprocess

Uses fetch to POST JSON to the endpoint and logs the parsed response. Replace <content-type> and <token> placeholders.

```javascript
const options = {
  method: 'POST',
  headers: {'Content-Type': '<content-type>', Authorization: 'Bearer <token>'},
  body: JSON.stringify({model: 'music-cover', audio_url: 'https://example.com/song.mp3'})
};

fetch('https://api.minimax.io/v1/music_cover_preprocess', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));
```

--------------------------------

### JavaScript fetch GET to retrieve file

Source: https://platform.minimax.io/docs/api-reference/file-management-retrieve

Uses fetch to send a GET request to the retrieve file endpoint and logs the JSON response. Replace <token> with your API token.

```javascript
const options = {method: 'GET', headers: {Authorization: 'Bearer <token>'}};

fetch('https://api.minimax.io/v1/files/retrieve', options)
  .then(res => res.json())
  .then(res => console.log(res))
  .catch(err => console.error(err));
```

--------------------------------

### Configure MiniMax in OpenHands

Source: https://platform.minimax.io/docs/m-plan/other-tools

Run a one-shot configuration command to set up MiniMax as the LLM provider in OpenHands. Replace `sk-cp-...` with your actual Subscription Key. This sets the model to `openai/MiniMax-M3.1-Flash` and the base URL to `https://api.minimax.io/v1`.

```bash
pipx run --spec openhands python -c '
import sys
from openhands_cli.stores.agent_store import AgentStore
AgentStore().create_and_save_from_settings(
    llm_api_key=sys.argv[1],
    settings={"llm_model": "openai/MiniMax-M3.1-Flash",
              "llm_base_url": "https://api.minimax.io/v1"},
)
' sk-cp-...
```

--------------------------------

### Python requests GET to retrieve file

Source: https://platform.minimax.io/docs/api-reference/file-management-retrieve

Uses the requests library to send a GET request to the retrieve file endpoint. Prints the response text. Replace <token> with your API token.

```python
import requests

url = "https://api.minimax.io/v1/files/retrieve"

headers = {"Authorization": "Bearer <token>"}

response = requests.get(url, headers=headers)

print(response.text)
```

--------------------------------

### First Request - Establish Cache (Anthropic SDK)

Source: https://platform.minimax.io/docs/api-reference/text-prompt-caching

Send the first request with the full content to create a cache entry. The response includes cache_read_input_tokens to show cache usage.

```python
  import anthropic

  client = anthropic.Anthropic()

response1 = client.messages.create(
      model="MiniMax-M3.1-Flash",
    system="You are an AI assistant tasked with analyzing literary works. Your goal is to provide insightful commentary on themes, characters, and writing style.\n",
      messages=[
          {
              "role": "user",
              "content": [
                  {
                      "type": "text",
                    "text": "<the entire contents of 'Pride and Prejudice'>"
                }
            ]
        },
    ],
    max_tokens=10240,
)

print("First request result:")
for block in response1.content:
    if block.type == "thinking":
        print(f"Thinking:\n{block.thinking}\n")
    elif block.type == "text":
        print(f"Output:\n{block.text}\n")
print(f"Input Tokens: {response1.usage.input_tokens}")
print(f"Output Tokens: {response1.usage.output_tokens}")
print(f"Cache Hit Tokens: {response1.usage.cache_read_input_tokens}")
```

--------------------------------

### Ruby Net::HTTP GET to retrieve file

Source: https://platform.minimax.io/docs/api-reference/file-management-retrieve

Uses Net::HTTP to send a GET request to the retrieve file endpoint and prints the response body. Replace <token> with your API token.

```ruby
require 'uri'
require 'net/http'

url = URI("https://api.minimax.io/v1/files/retrieve")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Get.new(url)
request["Authorization"] = 'Bearer <token>'

response = http.request(request)
puts response.read_body
```

--------------------------------

### Ruby Net::HTTP example for music_cover_preprocess

Source: https://platform.minimax.io/docs/api-reference/music-cover-preprocess

Uses Net::HTTP to POST JSON to the endpoint and prints the response body. Requires Ruby's standard library.

```ruby
require 'uri'
require 'net/http'

url = URI("https://api.minimax.io/v1/music_cover_preprocess")

http = Net::HTTP.new(url.host, url.port)
http.use_ssl = true

request = Net::HTTP::Post.new(url)
request["Content-Type"] = '<content-type>'
request["Authorization"] = 'Bearer <token>'
request.body = "{\n  \"model\": \"music-cover\",\n  \"audio_url\": \"https://example.com/song.mp3\"\n}"

response = http.request(request)
puts response.read_body
```

--------------------------------

### Task Start Event JSON payload

Source: https://platform.minimax.io/docs/api-reference/speech-t2a-websocket

Sent to the server to begin the speech synthesis task. The task is considered started only after receiving the `task_started` event. After that, you can send `task_continue` or `task_finish` events.

```json
{
  "event": "task_start",
  "model": "speech-2.8-turbo",
  "language_boost": "Chinese",
  "voice_setting": {
    "voice_id": "English_expressive_narrator",
    "speed": 1,
    "vol": 1,
    "pitch": 0
  },
  "pronunciation_dict": {
    "tone": [
      "Omg/Oh my god"
    ]
  },
  "audio_setting": {
    "sample_rate": 32000,
    "bitrate": 128000,
    "format": "mp3",
    "channel": 1
  }
}
```

--------------------------------

### PHP cURL GET to retrieve file

Source: https://platform.minimax.io/docs/api-reference/file-management-retrieve

Uses PHP cURL to send a GET request to the retrieve file endpoint. Prints the response or a cURL error if one occurs. Replace <token> with your API token.

```php
<?php

$curl = curl_init();

curl_setopt_array($curl, [
  CURLOPT_URL => "https://api.minimax.io/v1/files/retrieve",
  CURLOPT_RETURNTRANSFER => true,
  CURLOPT_ENCODING => "",
  CURLOPT_MAXREDIRS => 10,
  CURLOPT_TIMEOUT => 30,
  CURLOPT_HTTP_VERSION => CURL_HTTP_VERSION_1_1,
  CURLOPT_CUSTOMREQUEST => "GET",
  CURLOPT_HTTPHEADER => [
    "Authorization: Bearer <token>"
  ],
]);

$response = curl_exec($curl);
$err = curl_error($curl);

curl_close($curl);

if ($err) {
  echo "cURL Error #:" . $err;
} else {
  echo $response;
}
```

--------------------------------

### GET /v1/models

Source: https://platform.minimax.io/docs/api-reference/models/openai/list-models

Retrieves a list of available models. This endpoint returns an array of model objects with their identifiers and metadata.

```APIDOC
## [GET] /v1/models

### Description
Retrieves a list of available models. This endpoint returns an array of model objects with their identifiers and metadata.

### Method
GET

### Endpoint
https://api.minimax.io/v1/models

### Parameters
#### Headers
- **Authorization** (string) - Required - Bearer token for authentication.

### Request Example
```
curl --request GET \
  --url https://api.minimax.io/v1/models \
  --header 'Authorization: Bearer <token>'
```

### Response
#### Success Response (200)
- **object** (string) - The type of the response object, e.g., "list".
- **data** (array) - An array of model objects.
  - **id** (string) - The model identifier.
  - **object** (string) - The type of the object, e.g., "model".
  - **created** (integer) - Unix timestamp of model creation.
  - **owned_by** (string) - The owner of the model.

#### Response Example
```json
{
  "object": "list",
  "data": [
    {
      "id": "MiniMax-M3",
      "object": "model",
      "created": 1780272000,
      "owned_by": "minimax"
    },
    {
      "id": "MiniMax-M2.7",
      "object": "model",
      "created": 1773799200,
      "owned_by": "minimax"
    },
    {
      "id": "MiniMax-M2.5",
      "object": "model",
      "created": 1770948000,
      "owned_by": "minimax"
    }
  ]
}
```
```

--------------------------------

### Set Grok CLI env vars and launch with MiniMax model

Source: https://platform.minimax.io/docs/m-plan/other-tools

Sets GROK_BASE_URL and GROK_API_KEY environment variables, then launches Grok CLI with the MiniMax-M3.1-Flash model. Replace sk-cp-... with your M Plan API key.

```bash
export GROK_BASE_URL=https://api.minimax.io/v1
export GROK_API_KEY=sk-cp-...   # from M Plan
grok --model MiniMax-M3.1-Flash
```

--------------------------------

### Run wizard non-interactively with options

Source: https://platform.minimax.io/docs/m-plan/minimax-cli

Runs the wizard non-interactively for scripting. Only writes configuration and does not install tools. Must specify tools, key, and service region. Use --all to configure every supported tool and --dry-run to preview changes without network requests or file writes.

```bash
npx -y mmx-cli@latest agent setup \
  --agent claude-code \
  --agent codex \
  --api-key "$MINIMAX_API_KEY" \
  --region global
```
Tokens:
Raw