> Discover all available pages from the documentation index: https://mastra.zisheng.pro/zh-HK/llms.txt

# ![Friendli 標誌](https://models.dev/logos/friendli.svg)Friendli

透過 Mastra 的模型路由器存取 5 個 Friendli 模型。系統會自動使用 `FRIENDLI_TOKEN` 環境變數進行驗證。

詳情請參閱[Friendli 文件](https://friendli.ai/docs/guides/serverless_endpoints/introduction)。

```bash
FRIENDLI_TOKEN=your-api-token
```

```typescript
import { Agent } from '@mastra/core/agent'

const agent = new Agent({
  id: 'my-agent',
  name: 'My Agent',
  instructions: 'You are a helpful assistant',
  model: 'friendli/MiniMaxAI/MiniMax-M2.5',
})

// Generate a response
const response = await agent.generate('Hello!')

// Stream a response
const stream = await agent.stream('Tell me a story')
for await (const chunk of stream) {
  console.log(chunk)
}
```

> **資訊:** Mastra 使用與 OpenAI 相容的 `/chat/completions` 端點。部分 Provider 專屬功能可能無法使用，詳情請參閱[Friendli 文件](https://friendli.ai/docs/guides/serverless_endpoints/introduction)。

## 模型

| Model                                | Context | Tools | Reasoning | Image | Audio | Video | Input $/1M | Output $/1M |
| ------------------------------------ | ------- | ----- | --------- | ----- | ----- | ----- | ---------- | ----------- |
| `friendli/deepseek-ai/DeepSeek-V3.2` | 164K    |       |           |       |       |       | $0.50      | $2          |
| `friendli/google/gemma-4-31B-it`     | 262K    |       |           |       |       |       | $0.14      | $0.40       |
| `friendli/MiniMaxAI/MiniMax-M2.5`    | 197K    |       |           |       |       |       | $0.30      | $1          |
| `friendli/zai-org/GLM-5.1`           | 203K    |       |           |       |       |       | $1         | $4          |
| `friendli/zai-org/GLM-5.2`           | 1.0M    |       |           |       |       |       | $1         | $4          |

## 進階設定

### 自訂標頭

```typescript
const agent = new Agent({
  id: 'custom-agent',
  name: 'custom-agent',
  model: {
    url: 'https://api.friendli.ai/serverless/v1',
    id: 'friendli/MiniMaxAI/MiniMax-M2.5',
    apiKey: process.env.FRIENDLI_TOKEN,
    headers: {
      'X-Custom-Header': 'value',
    },
  },
})
```

### 動態選擇模型

```typescript
const agent = new Agent({
  id: 'dynamic-agent',
  name: 'Dynamic Agent',
  model: ({ requestContext }) => {
    const useAdvanced = requestContext.task === 'complex'
    return useAdvanced ? 'friendli/zai-org/GLM-5.2' : 'friendli/MiniMaxAI/MiniMax-M2.5'
  },
})
```