AWS Bedrock Proxy

A proxy that provides an OpenAI-compatible API interface for AWS Bedrock's models. Originally developed for use as a model provider in Xcode via Intelligence settings.

Features

• OpenAI-compatible API: Drop-in replacement for OpenAI's chat completions API
• Dynamic model discovery: Automatically loads available models from AWS Bedrock inference profiles
• Streaming support: Real-time streaming responses using Server-Sent Events (SSE)
• Authentication: Secure API key-based authentication
• Local development: Built-in HTTP server for local testing
• Multi-model support: Access all available models through a unified API

Prerequisites

• Node.js 18+ (for local development)
• AWS Account with Bedrock access
• AWS credentials configured

Installation

# Clone the repository
git clone git@github.com:grotter/bedrockproxy.git
cd bedrockproxy

# Install dependencies
npm install

Configuration

Environment Variables

Create a .env file for local development (see .env.template):

API_KEY=your-secret-api-key-here
MODEL_ID=us.anthropic.claude-sonnet-4-6
PORT=3000

• API_KEY: Secret key for authenticating API requests
• MODEL_ID: Default Claude model to use (optional, defaults to Sonnet 4.6)
• PORT: Local server port (optional, defaults to 3000)

Update Inference Profiles

To refresh the list of available models from AWS Bedrock:

aws bedrock list-inference-profiles --no-paginate > inference-profiles.json

Local Development

Start the local development server:

npm start

The server will start at http://localhost:3000 (or your configured PORT).

Testing Endpoints

List available models:

curl http://localhost:3000/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

Chat completion (non-streaming):

curl http://localhost:3000/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "us.anthropic.claude-sonnet-4-6",
    "messages": [
      {"role": "user", "content": "Hello!"}
    ],
    "max_tokens": 256
  }'

Chat completion (streaming):

curl http://localhost:3000/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "us.anthropic.claude-sonnet-4-6",
    "messages": [
      {"role": "user", "content": "Hello!"}
    ],
    "max_tokens": 256,
    "stream": true
  }'

API Reference

Authentication

All requests require authentication using either:

• Bearer token: Authorization: Bearer YOUR_API_KEY
• API key header: x-api-key: YOUR_API_KEY

Endpoints

GET /v1/models

Lists all available models

Response:

{
  "object": "list",
  "data": [
    {
      "id": "us.anthropic.claude-sonnet-4-6",
      "object": "model",
      "created": 1677610602,
      "owned_by": "anthropic"
    }
  ]
}

POST /v1/chat/completions

Creates a chat completion

Request body:

{
  "model": "us.anthropic.claude-sonnet-4-6",
  "messages": [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello!"}
  ],
  "max_tokens": 256,
  "stream": false
}

Parameters:

• model (string, optional): Model ID to use (defaults to MODEL_ID env var)
• messages (array, required): Array of message objects with role and content
• max_tokens (number, optional): Maximum tokens to generate (default: 256)
• stream (boolean, optional): Enable streaming responses (default: false)

Troubleshooting

Query string issues

The proxy automatically strips query strings from paths. Requests like /v1/models? will work correctly.

Authentication errors

Ensure your API_KEY environment variable is set and matches the key in your requests.

Model not found

Run aws bedrock list-inference-profiles to update your available models list.