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

# Introduction

> TokenDog — the unified LLM API gateway. One base URL for many models.

TokenDog is **the unified LLM API gateway**. Point your base URL at `https://tokendog.io`
to call OpenAI, Claude, Gemini, **plus Zhipu GLM and MiniMax** chat models, all through one
interface — reusing each vendor's existing clients and ecosystem.

<CardGroup cols={3}>
  <Card title="OpenAI-compatible" href="/en/api-reference/examples/basic-chat">
    Reuse the OpenAI SDK; paths like <code>/v1/chat/completions</code>, <code>/v1/responses</code>.
    Zhipu GLM and MiniMax models also use this format.
  </Card>

  <Card title="Claude Messages" href="/en/api-reference/examples/tool-calling">
    Native <code>/v1/messages</code> with tool use and multimodal input.
  </Card>

  <Card title="Gemini" href="/en/api-reference/examples/multimodal">
    Native <code>/v1beta/models/{'{model}'}:generateContent</code>.
  </Card>
</CardGroup>

## Supported models

One base URL and one key reach many providers' chat models — append the endpoint for the right format and set `model` in your request:

* **OpenAI-compatible** (`/v1/chat/completions`, `/v1/responses`): OpenAI (e.g. `gpt-5`, `gpt-5.1`), plus **Zhipu GLM (`glm-5.1`) and MiniMax (`MiniMax-M2.7`)**.
* **Claude Messages** (`/v1/messages`): Anthropic Claude (e.g. `claude-sonnet-4-6`).
* **Gemini** (`/v1beta/models/{'{model}'}:generateContent`): Google Gemini (e.g. `gemini-2.5-flash`).

<Note>For the complete, up-to-date model list, see the Model Marketplace in the console.</Note>

## What it solves

It consolidates many vendor APIs behind **one base URL and one key**: unified auth, billing,
and rate limiting, so clients don't integrate each vendor separately.

## Next

<Card title="Quickstart" href="/en/quickstart/quickstart">
  Send your first request in three steps.
</Card>