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

# Authentication

> Authenticate requests to the Portkey AI Gateway

## Overview

The Portkey AI Gateway uses provider-specific headers for authentication. You provide your API keys for the AI providers you want to use via HTTP headers.

## Provider Headers

Authentication is handled through HTTP headers that specify the provider and API key:

### Basic Authentication

<ParamField header="x-portkey-provider" type="string" required>
  The AI provider to use (e.g., `openai`, `anthropic`, `azure-openai`, `google`)
</ParamField>

<ParamField header="x-portkey-api-key" type="string" required>
  Your API key for the specified provider
</ParamField>

### Example: OpenAI

```bash theme={null}
curl http://localhost:8787/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "x-portkey-provider: openai" \
  -H "x-portkey-api-key: sk-..." \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

### Example: Anthropic

```bash theme={null}
curl http://localhost:8787/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "x-portkey-provider: anthropic" \
  -H "x-portkey-api-key: sk-ant-..." \
  -d '{
    "model": "claude-3-5-sonnet-20241022",
    "messages": [{"role": "user", "content": "Hello!"}],
    "max_tokens": 1024
  }'
```

## Azure OpenAI Headers

Azure OpenAI requires additional headers:

<ParamField header="x-portkey-provider" type="string" required>
  Set to `azure-openai`
</ParamField>

<ParamField header="x-portkey-api-key" type="string" required>
  Your Azure OpenAI API key
</ParamField>

<ParamField header="x-portkey-azure-resource-name" type="string" required>
  Your Azure resource name
</ParamField>

<ParamField header="x-portkey-azure-deployment-id" type="string" required>
  Your Azure deployment ID
</ParamField>

<ParamField header="x-portkey-azure-api-version" type="string">
  API version (defaults to latest)
</ParamField>

```bash theme={null}
curl http://localhost:8787/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "x-portkey-provider: azure-openai" \
  -H "x-portkey-api-key: your-azure-key" \
  -H "x-portkey-azure-resource-name: your-resource" \
  -H "x-portkey-azure-deployment-id: your-deployment" \
  -d '{
    "model": "gpt-4",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

## Config-Based Authentication

You can also pass authentication via the config header:

<ParamField header="x-portkey-config" type="string">
  JSON string containing provider and authentication details
</ParamField>

```bash theme={null}
curl http://localhost:8787/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H 'x-portkey-config: {"provider":"openai","api_key":"sk-..."}' \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

## Virtual Keys

<Note>
  Virtual keys are available in the hosted and enterprise versions of Portkey.
</Note>

Instead of passing provider API keys directly, you can use virtual keys:

```bash theme={null}
curl http://localhost:8787/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "x-portkey-virtual-key: pk-..." \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
```

## Supported Providers

The gateway supports authentication for 45+ providers including:

* **openai** - OpenAI
* **anthropic** - Anthropic
* **azure-openai** - Azure OpenAI
* **google** - Google Gemini
* **bedrock** - AWS Bedrock
* **cohere** - Cohere
* **mistral** - Mistral AI
* **together-ai** - Together AI
* **anyscale** - Anyscale
* **perplexity** - Perplexity
* **groq** - Groq
* **ollama** - Ollama

See the [Providers documentation](/providers/overview) for the complete list.