> ## Documentation Index
> Fetch the complete documentation index at: https://test-8862363a-tembo-byok-agent-subscription-updates.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Max

> One subscription, all coding agents and models.

<Frame>
  <img src="https://mintlify.s3.us-west-1.amazonaws.com/test-8862363a-tembo-byok-agent-subscription-updates/images/max-light.png" alt="Tembo Max" className="block dark:hidden" />

  <img src="https://mintlify.s3.us-west-1.amazonaws.com/test-8862363a-tembo-byok-agent-subscription-updates/images/max-dark.png" alt="Tembo Max" className="hidden dark:block" />
</Frame>

Tembo Max gives you access to all coding agents and models through a single subscription. Connect your local tools to Tembo's proxy with one API key, no separate provider accounts needed.

Available on all plans, including the free tier. [Read more about Tembo Max](https://tembo.io/max).

BYOK is also available on all plans from **Settings -> Models -> API Keys**. Paid Tembo plans can connect supported agent subscriptions, including Codex, from **Settings -> Connected accounts -> LLM subscriptions**.

## Setup

1. Get your API key from **Settings > API Keys** in the [dashboard](https://app.tembo.io)
2. Point your tool at `proxy.tembo.io`

<CodeGroup>
  ```bash Claude Code theme={null}
  export ANTHROPIC_BASE_URL=https://proxy.tembo.io/anthropic
  export ANTHROPIC_API_KEY=your-tembo-api-key
  ```

  ```bash Codex theme={null}
  export OPENAI_BASE_URL="https://proxy.tembo.io/openai"
  export OPENAI_API_KEY="your-tembo-api-key"
  ```

  ```json OpenCode theme={null}
  {
    "$schema": "https://opencode.ai/config.json",
    "model": "openai/gpt-5-mini",
    "provider": {
      "openai": {
        "models": {
          "gpt-5.4": {
            "id": "gpt-5.4",
            "name": "GPT-5.4"
          },
          "gpt-5.4-mini": {
            "id": "gpt-5-mini",
            "name": "GPT-5.4 Mini"
          },
          "gpt-5.4-nano": {
            "id": "gpt-5-nano",
            "name": "GPT-5.4 Nano"
          }
        },
        "options": {
          "baseURL": "https://proxy.tembo.io/openai",
          "apiKey": "<your-tembo-api-key>",
          "headers": {
            "x-reasoning-level": "high"
          }
        }
      }
    }
  }
  ```
</CodeGroup>

<Note>
  Codex may store credentials in `~/.codex/auth.json` that override environment variables.
</Note>

Codex and OpenCode both support `gpt-5.4`, `gpt-5.4-mini`, and `gpt-5.4-nano`. In OpenCode, the `gpt-5.4-mini` and `gpt-5.4-nano` aliases map to OpenAI's `gpt-5-mini` and `gpt-5-nano` model IDs.

## Proxy models

All models are tested across real-world coding work before being made available.

| Model             | Model ID                   | Endpoint                   |
| ----------------- | -------------------------- | -------------------------- |
| Claude Opus 4.5   | `claude-opus-4-5`          | `proxy.tembo.io/anthropic` |
| Claude 4.5 Sonnet | `claude-4-5-sonnet`        | `proxy.tembo.io/anthropic` |
| Claude 4.1 Opus   | `claude-4.1-opus`          | `proxy.tembo.io/anthropic` |
| Claude 4.5 Haiku  | `claude-4-5-haiku`         | `proxy.tembo.io/anthropic` |
| Claude 4 Sonnet   | `claude-4-sonnet`          | `proxy.tembo.io/anthropic` |
| GPT-5.4           | `gpt-5.4`                  | `proxy.tembo.io/openai`    |
| GPT-5.4 Mini      | `gpt-5.4-mini`             | `proxy.tembo.io/openai`    |
| GPT-5.4 Nano      | `gpt-5.4-nano`             | `proxy.tembo.io/openai`    |
| GPT-5.2           | `gpt-5.2`                  | `proxy.tembo.io/openai`    |
| GPT-5.1 Codex     | `gpt-5.1-codex`            | `proxy.tembo.io/openai`    |
| GPT-5.1 Codex Max | `gpt-5.1-codex-max`        | `proxy.tembo.io/openai`    |
| GPT-5.2 Codex     | `gpt-5.2-codex`            | `proxy.tembo.io/openai`    |
| Gemini 2.5 Pro    | `gemini-2.5-pro`           | `proxy.tembo.io/anthropic` |
| Gemini 3 Pro      | `gemini-3-pro`             | `proxy.tembo.io/anthropic` |
| Gemini 3 Flash    | `gemini-3-flash`           | `proxy.tembo.io/anthropic` |
| Kimi K2           | `bedrock-kimi-k2-thinking` | `proxy.tembo.io/anthropic` |
| Grok              | `grok`                     | Cursor API                 |

Tembo-managed Anthropic and Google models require no additional provider keys. Some OpenAI models require BYOK with your own OpenAI API key, and paid Tembo plans can use supported agent subscriptions such as Codex where available.

## Pricing

See [tembo.io/pricing](https://tembo.io/pricing) for plans and credits.

## Privacy

| Provider  | Data Retention   |
| --------- | ---------------- |
| Anthropic | Zero retention   |
| Google    | Zero retention   |
| OpenAI    | 30-day retention |