labring · octo-patch · Mar 14, 2026
diff --git a/document/content/docs/self-host/config/model/meta.en.json b/document/content/docs/self-host/config/model/meta.en.json
@@ -1,4 +1,4 @@
 {
   "title": "Model Configuration",
-  "pages": ["intro", "ai-proxy", "one-api", "siliconCloud", "ppio"]
+  "pages": ["intro", "ai-proxy", "one-api", "siliconCloud", "ppio", "minimax"]
 }
diff --git a/document/content/docs/self-host/config/model/meta.json b/document/content/docs/self-host/config/model/meta.json
@@ -1,4 +1,4 @@
 {
   "title": "模型配置方案",
-  "pages": ["intro", "ai-proxy", "one-api", "siliconCloud", "ppio"]
+  "pages": ["intro", "ai-proxy", "one-api", "siliconCloud", "ppio", "minimax"]
 }
diff --git a/document/content/docs/self-host/config/model/minimax.en.mdx b/document/content/docs/self-host/config/model/minimax.en.mdx
@@ -0,0 +1,106 @@
+---
+title: Connect Models via MiniMax
+description: Connect Models via MiniMax
+---
+
+import { Alert } from '@/components/docs/Alert';
+
+[MiniMax](https://www.minimaxi.com) is an AI technology company that provides high-performance large language model API services. MiniMax's API is compatible with the OpenAI format, making it easy to integrate with FastGPT.
+
+## 1. Get a MiniMax API Key
+
+1. Visit [MiniMax Platform](https://platform.minimaxi.com), register and log in.
+2. Go to the console and create an API Key.
+
+## 2. Configure Models
+
+### Option 1: Via OneAPI
+
+Refer to the [OneAPI Integration Guide](/docs/self-host/config/model/one-api) to add a MiniMax channel in OneAPI:
+
+- Channel type: Select **MiniMax** or **Custom Channel**
+- Base URL: `https://api.minimax.io/v1`
+- Enter your MiniMax API Key
+- Model: `MiniMax-M2.5` or `MiniMax-M2.5-highspeed`
+
+Once configured, enable the corresponding model in FastGPT.
+
+### Option 2: Direct Integration (without OneAPI)
+
+In the FastGPT model configuration page, add a custom model with the following information:
+
+- **Model ID**: `MiniMax-M2.5`
+- **Custom Request URL**: `https://api.minimax.io/v1/chat/completions`
+- **Custom Request Key**: Enter your MiniMax API Key
+
+<Alert icon="⚠️" context="warning">
+  MiniMax models require temperature to be in the range (0.0, 1.0]. A value of 0 is not supported. It is recommended to set the maximum temperature to 1.0.
+</Alert>
+
+### Via Configuration File
+
+If you prefer configuring models via a configuration file, refer to the following JSON:
+
+```json
+[
+  {
+    "model": "MiniMax-M2.5",
+    "metadata": {
+      "isCustom": true,
+      "isActive": true,
+      "provider": "MiniMax",
+      "model": "MiniMax-M2.5",
+      "name": "MiniMax-M2.5",
+      "maxContext": 204000,
+      "maxResponse": 16000,
+      "quoteMaxToken": 200000,
+      "maxTemperature": 1.0,
+      "vision": true,
+      "datasetProcess": true,
+      "usedInClassify": true,
+      "usedInExtractFields": true,
+      "usedInToolCall": true,
+      "toolChoice": true,
+      "functionCall": true,
+      "defaultSystemChatPrompt": "",
+      "defaultConfig": {},
+      "requestUrl": "https://api.minimax.io/v1/chat/completions",
+      "requestAuth": "your-minimax-api-key"
+    }
+  },
+  {
+    "model": "MiniMax-M2.5-highspeed",
+    "metadata": {
+      "isCustom": true,
+      "isActive": true,
+      "provider": "MiniMax",
+      "model": "MiniMax-M2.5-highspeed",
+      "name": "MiniMax-M2.5-highspeed",
+      "maxContext": 204000,
+      "maxResponse": 16000,
+      "quoteMaxToken": 200000,
+      "maxTemperature": 1.0,
+      "vision": true,
+      "datasetProcess": true,
+      "usedInClassify": true,
+      "usedInExtractFields": true,
+      "usedInToolCall": true,
+      "toolChoice": true,
+      "functionCall": true,
+      "defaultSystemChatPrompt": "",
+      "defaultConfig": {}
+    }
+  }
+]
+```
+
+## 3. Available Models
+
+| Model Name | Max Context | Features |
+| --- | --- | --- |
+| MiniMax-M2.5 | 204K tokens | High-performance general model with vision and function calling support |
+| MiniMax-M2.5-highspeed | 204K tokens | High-speed version, ideal for latency-sensitive scenarios |
+
+## 4. Testing
+
+After configuration, click the test button on the FastGPT model configuration page to verify the model is working properly.
diff --git a/document/content/docs/self-host/config/model/minimax.mdx b/document/content/docs/self-host/config/model/minimax.mdx
@@ -0,0 +1,106 @@
+---
+title: 通过 MiniMax 接入模型
+description: 通过 MiniMax 接入模型
+---
+
+import { Alert } from '@/components/docs/Alert';
+
+[MiniMax](https://www.minimaxi.com) 是一家通用人工智能科技公司，提供高性能的大语言模型 API 服务。MiniMax 的 API 兼容 OpenAI 格式，可以方便地接入 FastGPT。
+
+## 1. 获取 MiniMax API Key
+
+1. 访问 [MiniMax 开放平台](https://platform.minimaxi.com)，注册并登录账号。
+2. 进入控制台，创建 API Key。
+
+## 2. 配置模型
+
+### 方式一：通过 OneAPI 接入
+
+参考 [OneAPI 接入教程](/docs/self-host/config/model/one-api)，在 OneAPI 中添加 MiniMax 渠道：
+
+- 渠道类型选择 **MiniMax** 或 **自定义渠道**
+- Base URL 填写：`https://api.minimax.io/v1`
+- 填写 MiniMax 的 API Key
+- 模型填写：`MiniMax-M2.5` 或 `MiniMax-M2.5-highspeed`
+
+配置完成后，在 FastGPT 中启用对应模型即可。
+
+### 方式二：直接接入（不使用 OneAPI）
+
+在 FastGPT 模型配置页面，添加自定义模型，填写以下信息：
+
+- **模型 ID**：`MiniMax-M2.5`
+- **自定义请求地址**：`https://api.minimax.io/v1/chat/completions`
+- **自定义请求 Key**：填写 MiniMax 的 API Key
+
+<Alert icon="⚠️" context="warning">
+  MiniMax 模型的 temperature 参数范围为 (0.0, 1.0]，不支持设置为 0。建议将最大温度设置为 1.0。
+</Alert>
+
+### 通过配置文件接入
+
+如果你更习惯通过配置文件配置模型，可以参考以下 JSON 配置：
+
+```json
+[
+  {
+    "model": "MiniMax-M2.5",
+    "metadata": {
+      "isCustom": true,
+      "isActive": true,
+      "provider": "MiniMax",
+      "model": "MiniMax-M2.5",
+      "name": "MiniMax-M2.5",
+      "maxContext": 204000,
+      "maxResponse": 16000,
+      "quoteMaxToken": 200000,
+      "maxTemperature": 1.0,
+      "vision": true,
+      "datasetProcess": true,
+      "usedInClassify": true,
+      "usedInExtractFields": true,
+      "usedInToolCall": true,
+      "toolChoice": true,
+      "functionCall": true,
+      "defaultSystemChatPrompt": "",
+      "defaultConfig": {},
+      "requestUrl": "https://api.minimax.io/v1/chat/completions",
+      "requestAuth": "your-minimax-api-key"
+    }
+  },
+  {
+    "model": "MiniMax-M2.5-highspeed",
+    "metadata": {
+      "isCustom": true,
+      "isActive": true,
+      "provider": "MiniMax",
+      "model": "MiniMax-M2.5-highspeed",
+      "name": "MiniMax-M2.5-highspeed",
+      "maxContext": 204000,
+      "maxResponse": 16000,
+      "quoteMaxToken": 200000,
+      "maxTemperature": 1.0,
+      "vision": true,
+      "datasetProcess": true,
+      "usedInClassify": true,
+      "usedInExtractFields": true,
+      "usedInToolCall": true,
+      "toolChoice": true,
+      "functionCall": true,
+      "defaultSystemChatPrompt": "",
+      "defaultConfig": {}
+    }
+  }
+]
+```
+
+## 3. 可用模型
+
+| 模型名称 | 最大上下文 | 特点 |
+| --- | --- | --- |
+| MiniMax-M2.5 | 204K tokens | 高性能通用模型，支持视觉、函数调用 |
+| MiniMax-M2.5-highspeed | 204K tokens | 高速版本，适合对响应速度要求较高的场景 |
+
+## 4. 测试
+
+配置完成后，可以在 FastGPT 模型配置页面点击测试按钮，验证模型是否正常工作。