Bedrock's models, on a portable key.
Keep Bedrock for AWS-governed workloads. Outside that boundary, the same Claude, Llama, and Mistral families pass through ProxyLLM on your own OpenRouter key, behind sub-keys, budget caps, and request logs.
$129/month SaaS. Bring your own model keys. No inference markup.
Three steps to connect.
Know the Bedrock boundary
Amazon Bedrock uses AWS auth and provider-specific model IDs. ProxyLLM does not claim native Bedrock signing or key storage today.
Pass the same model families through
Claude, Llama, and Mistral also ship through OpenRouter. Add your own OpenRouter key and call them at https://api.proxyllm.ai/v1 from the OpenAI SDK you already use.
Scope who spends what
Give each team a scoped sub-key with its own budget cap instead of shared provider credentials, then read the request logs to see what every workload costs.
Model equivalents, not fake AWS auth.
ProxyLLM passes Claude, Llama, and Mistral traffic through your OpenRouter key today. Native Bedrock signing is a future adapter.
from openai import OpenAI
client = OpenAI(
base_url="https://api.proxyllm.ai/v1",
api_key="pk_live_...",
)
r = client.chat.completions.create(
model="anthropic/claude-3.5-haiku",
messages=[{"role": "user", "content": "Run the Bedrock-equivalent fallback."}],
)Run your AI workloads on your ChatGPT subscription.
ProxyLLM runs OpenAI's Codex for you, signed in with your own ChatGPT account. Your apps call one OpenAI-compatible endpoint and the work bills to your flat plan instead of per-token API pricing.
Keep Bedrock from becoming a silo.
Run the same model families outside AWS on a key you control. ProxyLLM adds sub-keys, budget caps, and request logs, with no markup on inference.