Amazon Bedrock
Amazon Bedrock provides managed access to foundation models including Amazon's Nova family. AgentXchain connects via api_proxy using Bedrock's OpenAI-compatible inference endpoint.
Which adapter?
api_proxy with provider: "openai" and a custom base_url pointing to your Bedrock endpoint.
Prerequisites
- An AWS account with Bedrock model access enabled
- Bedrock API credentials
agentxchainCLI installed
Configuration
{
"runtimes": {
"nova-dev": {
"type": "api_proxy",
"provider": "openai",
"model": "amazon.nova-2-pro",
"auth_env": "BEDROCK_API_KEY",
"base_url": "https://bedrock-runtime.us-east-1.amazonaws.com/model/amazon.nova-2-pro/converse"
}
},
"roles": {
"dev": {
"runtime": "nova-dev",
"mandate": "Implement features and fix bugs",
"authority": "proposed"
}
}
}
Available models
| Model | Best for |
|---|---|
amazon.nova-2-pro | Most capable Nova model |
amazon.nova-2-lite | Fast, cost-effective |
amazon.nova-premier | Premium reasoning |
Verify the connection
export BEDROCK_API_KEY="..."
agentxchain connector check
Gotchas
- Authentication: Bedrock uses AWS IAM credentials, not simple API keys. You may need a wrapper script that signs requests with AWS SigV4. The simplest approach is to use a local proxy that handles AWS auth and exposes an OpenAI-compatible endpoint.
- Region-specific endpoints: Bedrock endpoints are region-specific. Update
base_urlto match your region. - Model access: Nova models must be explicitly enabled in your AWS account via the Bedrock console before use.
- Cost rates: Supply operator-specific rates:
{
"budget": {
"cost_rates": {
"amazon.nova-2-pro": { "input_per_million": 0.80, "output_per_million": 3.20 }
}
}
}