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

# AI Gateway Overview

> Access 100+ AI models through one unified API with intelligent routing, automatic fallbacks, and unified observability

## What is Helicone AI Gateway?

Helicone AI Gateway is a unified inference platform that provides access to 100+ AI models from multiple providers through a single OpenAI-compatible API. Simply change your `baseURL` to start using models from OpenAI, Anthropic, Google, Meta, and dozens of other providers.

<CardGroup cols={2}>
  <Card title="Quick Start" icon="rocket" href="/gateway/getting-started">
    Get started in 2 minutes with one line of code
  </Card>

  <Card title="Routing" icon="route" href="/gateway/routing">
    Intelligent routing across providers
  </Card>

  <Card title="Fallbacks" icon="shield" href="/gateway/fallbacks">
    Automatic failover for reliability
  </Card>

  <Card title="Integrations" icon="plug" href="/gateway/integrations">
    Works with your existing tools
  </Card>
</CardGroup>

## Key Features

<AccordionGroup>
  <Accordion title="Unified API" icon="layer-group">
    Access models from 20+ providers using the OpenAI API format. No need to learn multiple SDKs or API formats.

    ```typescript theme={null}
    // Same API for all providers
    const response = await client.chat.completions.create({
      model: "claude-sonnet-4",  // or gpt-4o, gemini-2.0-flash, llama-3.3-70b
      messages: [{ role: "user", content: "Hello!" }]
    });
    ```
  </Accordion>

  <Accordion title="Intelligent Routing" icon="arrows-split-up-and-left">
    Automatically route requests to the best available provider based on:

    * Cost optimization
    * Latency requirements
    * Provider availability
    * Your BYOK (Bring Your Own Key) preferences

    Learn more in [Routing](/gateway/routing).
  </Accordion>

  <Accordion title="Automatic Fallbacks" icon="rotate">
    Built-in resilience with automatic failover when a provider fails:

    ```typescript theme={null}
    // Tries Bedrock first, falls back to Anthropic
    model: "claude-3-7-sonnet-20250219/bedrock,claude-3-7-sonnet-20250219/anthropic"
    ```

    Learn more in [Fallbacks](/gateway/fallbacks).
  </Accordion>

  <Accordion title="Cost Tracking" icon="dollar-sign">
    Unified cost tracking across all providers with accurate pricing for 300+ models.
  </Accordion>

  <Accordion title="Unified Observability" icon="chart-line">
    Single dashboard to monitor requests across all providers with traces, logs, and analytics.
  </Accordion>
</AccordionGroup>

## Supported Providers

The AI Gateway supports 20+ inference providers:

<CardGroup cols={3}>
  <Card title="OpenAI" icon="openai">
    GPT-4o, GPT-5, o1, o3, o4
  </Card>

  <Card title="Anthropic" icon="a">
    Claude 3.7 Sonnet, Claude 4
  </Card>

  <Card title="Google" icon="google">
    Gemini 2.0 Flash, Gemini 3
  </Card>

  <Card title="Meta" icon="meta">
    Llama 3.3, Llama 4
  </Card>

  <Card title="AWS Bedrock" icon="aws">
    All Bedrock models
  </Card>

  <Card title="Google Vertex" icon="google">
    Vertex AI models
  </Card>

  <Card title="DeepSeek" icon="code">
    DeepSeek models
  </Card>

  <Card title="Mistral" icon="wind">
    Mistral models
  </Card>

  <Card title="xAI" icon="x-twitter">
    Grok models
  </Card>

  <Card title="Groq" icon="bolt">
    Fast inference
  </Card>

  <Card title="DeepInfra" icon="server">
    100+ models
  </Card>

  <Card title="OpenRouter" icon="route">
    Multi-provider routing
  </Card>
</CardGroup>

<Info>
  View the complete list of supported models and providers at [helicone.ai/models](https://www.helicone.ai/models)
</Info>

## How It Works

<Steps>
  <Step title="Change Your Base URL">
    Update your API client to point to Helicone's AI Gateway:

    ```typescript theme={null}
    const client = new OpenAI({
      baseURL: "https://ai-gateway.helicone.ai",
      apiKey: process.env.HELICONE_API_KEY,
    });
    ```
  </Step>

  <Step title="Specify Models">
    Use model names directly or specify providers:

    ```typescript theme={null}
    // Auto-routing across providers
    model: "gpt-4o-mini"

    // Specific provider
    model: "claude-sonnet-4/anthropic"

    // With fallback
    model: "gpt-4o/openai,gpt-4o/azure"
    ```
  </Step>

  <Step title="Make Requests">
    Use the standard OpenAI API format for all requests:

    ```typescript theme={null}
    const response = await client.chat.completions.create({
      model: "claude-sonnet-4",
      messages: [{ role: "user", content: "Hello!" }]
    });
    ```
  </Step>
</Steps>

## Authentication Options

<Tabs>
  <Tab title="Pass-Through Billing (PTB)">
    Use Helicone's API key and pay through Helicone:

    ```typescript theme={null}
    const client = new OpenAI({
      baseURL: "https://ai-gateway.helicone.ai",
      apiKey: process.env.HELICONE_API_KEY,
    });
    ```

    **Benefits:**

    * Single API key for all providers
    * Consolidated billing
    * No provider account setup needed
    * Automatic credit management
  </Tab>

  <Tab title="Bring Your Own Key (BYOK)">
    Use your own provider API keys:

    ```typescript theme={null}
    const client = new OpenAI({
      baseURL: "https://ai-gateway.helicone.ai",
      apiKey: process.env.HELICONE_API_KEY,
    });
    ```

    Then add your provider keys in the [Helicone dashboard](https://helicone.ai/settings/keys).

    **Benefits:**

    * Use existing provider credits
    * Direct billing from providers
    * Priority support with BYOK attempts first
  </Tab>
</Tabs>

## Model Discovery

The AI Gateway provides an OpenAI-compatible `/v1/models` endpoint to discover available models:

```bash theme={null}
curl https://ai-gateway.helicone.ai/v1/models \
  -H "Authorization: Bearer $HELICONE_API_KEY"
```

<ResponseExample>
  ```json theme={null}
  {
    "object": "list",
    "data": [
      {
        "id": "gpt-4o-mini",
        "object": "model",
        "created": 1716537600,
        "owned_by": "openai"
      },
      {
        "id": "claude-sonnet-4",
        "object": "model",
        "created": 1708560000,
        "owned_by": "anthropic"
      }
    ]
  }
  ```
</ResponseExample>

## Next Steps

<CardGroup cols={2}>
  <Card title="Get Started" icon="play" href="/gateway/getting-started">
    Start using the AI Gateway in 2 minutes
  </Card>

  <Card title="Learn Routing" icon="map" href="/gateway/routing">
    Understand how requests are routed
  </Card>

  <Card title="Configure Fallbacks" icon="shield-check" href="/gateway/fallbacks">
    Set up automatic failover
  </Card>

  <Card title="Browse Models" icon="grid" href="https://www.helicone.ai/models">
    Explore 300+ available models
  </Card>
</CardGroup>