> ## Documentation Index
> Fetch the complete documentation index at: https://www.adaline.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Together AI

> Integrate Together AI models through the Adaline Proxy for automatic telemetry and observability.

# Together AI

Integrate Together AI open-source models through the Adaline Proxy to automatically capture telemetry — requests, responses, token usage, latency, and costs — with minimal code changes.

## Supported Models

Together AI accepts any model name available on the Together AI platform. Popular models include:

**Chat Models**

| Model                                      | Description                 |
| ------------------------------------------ | --------------------------- |
| `meta-llama/Llama-3.1-405B-Instruct-Turbo` | Llama 3.1 405B              |
| `meta-llama/Llama-3.1-70B-Instruct-Turbo`  | Llama 3.1 70B               |
| `meta-llama/Llama-3.1-8B-Instruct-Turbo`   | Llama 3.1 8B                |
| `mistralai/Mixtral-8x7B-Instruct-v0.1`     | Mixtral 8x7B MoE            |
| `Qwen/Qwen2-72B-Instruct`                  | Qwen 2 72B                  |
| `deepseek-ai/DeepSeek-R1`                  | DeepSeek R1 reasoning model |

**Embedding Models**

Together AI also supports embedding models. Pass any Together AI embedding model name.

<Info>
  Together AI uses a flexible model routing system — any model available on the [Together AI models page](https://api.together.xyz/models) can be used.
</Info>

## Proxy Base URL

```
https://gateway.adaline.ai/v1/together-ai/
```

## Prerequisites

1. A [Together AI API key](https://api.together.xyz/settings/api-keys)
2. An [Adaline API key](/docs/admin/create-api-keys), project ID, and prompt ID

## Chat Completions

### Complete Chat

<CodeGroup>
  ```python Python theme={null}
  from together import Together

  client = Together(
      base_url="https://gateway.adaline.ai/v1/together-ai",
      api_key="your-together-ai-api-key",
      supplied_headers={
          "adaline-api-key": "your-adaline-api-key",
          "adaline-project-id": "your-project-id",
          "adaline-prompt-id": "your-prompt-id",
      },
  )

  response = client.chat.completions.create(
      model="meta-llama/Llama-2-7b-chat-hf",
      messages=[
          {"role": "system", "content": "You are a helpful assistant."},
          {"role": "user", "content": "What are the benefits of open source AI models?"}
      ],
      extra_headers={
          "adaline-trace-name": "togetherai-chat-completion"  # Optional
      }
  )

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

  ```typescript TypeScript theme={null}
  import OpenAI from "openai";

  // Together AI is OpenAI-compatible
  const client = new OpenAI({
    apiKey: "your-together-ai-api-key",
    baseURL: "https://gateway.adaline.ai/v1/together-ai/",
  });

  const response = await client.chat.completions.create(
    {
      model: "meta-llama/Llama-2-7b-chat-hf",
      messages: [
        { role: "system", content: "You are a helpful assistant." },
        { role: "user", content: "What are the benefits of open source AI models?" },
      ],
    },
    {
      headers: {
        "adaline-api-key": "your-adaline-api-key",
        "adaline-project-id": "your-project-id",
        "adaline-prompt-id": "your-prompt-id",
        "adaline-trace-name": "togetherai-chat-completion",
      },
    }
  );

  console.log(response.choices[0].message.content);
  ```
</CodeGroup>

<Note>
  Together AI's Python SDK uses `supplied_headers` for persistent headers set at client initialization, and `extra_headers` for per-request headers.
</Note>

### Stream Chat

<CodeGroup>
  ```python Python theme={null}
  from together import Together

  client = Together(
      base_url="https://gateway.adaline.ai/v1/together-ai",
      api_key="your-together-ai-api-key",
      supplied_headers={
          "adaline-api-key": "your-adaline-api-key",
          "adaline-project-id": "your-project-id",
          "adaline-prompt-id": "your-prompt-id",
      },
  )

  stream = client.chat.completions.create(
      model="meta-llama/Llama-2-7b-chat-hf",
      messages=[
          {"role": "system", "content": "You are a helpful assistant."},
          {"role": "user", "content": "Explain the concept of distributed computing."}
      ],
      stream=True,
      extra_headers={
          "adaline-trace-name": "togetherai-stream-chat"  # Optional
      }
  )

  for chunk in stream:
      if chunk.choices[0].delta.content is not None:
          print(chunk.choices[0].delta.content, end="")
  ```
</CodeGroup>

## Next Steps

* [Multi-Step Workflows](/docs/integrations/examples/multi-step-workflows) — RAG pipelines, multi-step generation, and conversational agents
* [Headers Reference](/docs/reference/proxy/headers) — Complete header documentation

***

<Card title="Back to Integrations" icon="arrow-left" href="/docs/integrations/introduction">
  Browse all integrations
</Card>