> ## Documentation Index
> Fetch the complete documentation index at: https://tokonomics.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Supported AI Providers, Slugs, and Proxy Endpoints

> Complete list of LLM providers supported by Tokonomics, including their proxy slugs, upstream base paths, and compatible API formats.

Tokonomics proxies your requests to all major LLM providers through a single, unified endpoint. Each provider is identified by a short **slug** that you embed in the proxy URL. The table below lists every supported provider, its slug, the upstream base URL Tokonomics forwards to, the API format it uses, and a typical example path.

## Supported providers

| Provider      | Slug        | Upstream base URL                           | API format        | Example path                            |
| ------------- | ----------- | ------------------------------------------- | ----------------- | --------------------------------------- |
| OpenAI        | `openai`    | `https://api.openai.com/v1`                 | OpenAI            | `chat/completions`                      |
| Anthropic     | `anthropic` | `https://api.anthropic.com`                 | Anthropic         | `messages`                              |
| DeepSeek      | `deepseek`  | `https://api.deepseek.com/v1`               | OpenAI-compatible | `chat/completions`                      |
| Google Gemini | `gemini`    | `https://generativelanguage.googleapis.com` | Gemini            | `v1beta/models/{model}:generateContent` |
| Mistral       | `mistral`   | `https://api.mistral.ai/v1`                 | OpenAI-compatible | `chat/completions`                      |
| xAI (Grok)    | `xai`       | `https://api.x.ai/v1`                       | OpenAI-compatible | `chat/completions`                      |
| Cohere        | `cohere`    | `https://api.cohere.com/v1`                 | Cohere            | `chat`                                  |
| Groq          | `groq`      | `https://api.groq.com/openai/v1`            | OpenAI-compatible | `chat/completions`                      |
| Together AI   | `together`  | `https://api.together.xyz/v1`               | OpenAI-compatible | `chat/completions`                      |

## URL construction

Every Tokonomics proxy call follows this pattern:

```text theme={null}
https://tokonomics.ca/proxy/{slug}/{path}
```

* **`{slug}`** — the provider identifier from the table above (e.g. `openai`, `anthropic`)
* **`{path}`** — the upstream API path you would normally append to the provider's base URL, **without** repeating the base URL itself

For example, to call OpenAI's chat completions endpoint you would use:

```text theme={null}
https://tokonomics.ca/proxy/openai/chat/completions
```

And to call Anthropic's messages endpoint:

```text theme={null}
https://tokonomics.ca/proxy/anthropic/messages
```

Tokonomics strips the `https://tokonomics.ca/proxy/{slug}` prefix and reconstructs the full upstream URL automatically before forwarding your request.

## Authentication to providers

Every proxied request carries **two sets of credentials**:

1. **Your Tokonomics key** — passed in the standard `Authorization` header as a Bearer token. Tokonomics reads this to authenticate you, meter usage, and enforce budget caps.
2. **Your provider API key** — passed in whichever header the upstream provider requires (e.g. `Authorization` for OpenAI, `x-api-key` for Anthropic). Tokonomics forwards this header to the upstream provider unchanged.

A typical request to OpenAI through the proxy looks like this:

```bash theme={null}
curl https://tokonomics.ca/proxy/openai/chat/completions \
  -H "Authorization: Bearer mk_your_key" \
  -H "X-Provider-Authorization: Bearer sk_your_openai_key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

Tokonomics strips the `Authorization: Bearer mk_...` header before it reaches the upstream provider, and passes all other headers — including your provider key — through unchanged.

<Note>
  OpenAI-compatible providers (DeepSeek, Mistral, xAI, Groq, and Together AI) work out of the box with the official `openai` Python or Node.js SDK. Simply set `base_url` to the Tokonomics proxy URL for that provider — no other changes required.

  ```python theme={null}
  from openai import OpenAI

  client = OpenAI(
      api_key="sk_your_deepseek_key",
      base_url="https://tokonomics.ca/proxy/deepseek",
  )
  ```

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

  const client = new OpenAI({
    apiKey: "sk_your_groq_key",
    baseURL: "https://tokonomics.ca/proxy/groq",
  });
  ```
</Note>