Skip to main content

OpenAI SDK Compatibility

OpenProxyAI exposes an OpenAI-compatible API surface — /v1/chat/completions, /v1/embeddings, /v1/models. If your application already uses the official OpenAI SDK, switching to OpenProxyAI is a base_url change, not a rewrite.

from openai import OpenAI

client = OpenAI(
    base_url="https://api.openproxy.ai/v1",
    api_key=os.environ["OPENPROXY_KEY"],
)

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": "hi"}],
)

Same for the TypeScript/JavaScript openai package — pass baseURL in the client constructor.

Adding OpenProxyAI-specific headers​

Optional headers extend the request without breaking compatibility with the underlying OpenAI client:

x-op-team:   eng-platform
x-op-policy: pii_redact,topic_guard

See Rate Limiting & Budgets for x-op-team, and Policies & Guardrails for x-op-policy.

Calling non-OpenAI models​

The same OpenAI-shaped request works for any provider behind OpenProxyAI — set model to the target model name (for example claude-3-5-sonnet or gemini-1.5-pro) and OpenProxyAI routes it to the right provider. See Provider Routing.

Using the dedicated SDKs instead​

For richer error types and typed gateway metadata (cost, latency, policy decisions) on every response, see the Python SDK and TypeScript SDK.