An OpenAI-compatible API for every model

Router One exposes an OpenAI-compatible endpoint at https://api.router.one/v1. Change one base URL, keep the OpenAI SDK you already use, and get instant access to Claude, GPT, Gemini, Grok and 30+ models through a single endpoint — with smart routing, automatic provider fallback, and a per-request cost trace for every call.

Drop in by changing one base URL

Router One speaks the OpenAI Chat Completions API. Point your client at the Router One base URL, swap in your Router One key, and the rest of your code is unchanged — no new SDK, no rewrite.

# Keep your OpenAI SDK — just point it at Router One
export OPENAI_BASE_URL=https://api.router.one/v1
export OPENAI_API_KEY=sk-your-router-one-key

Call any model through the OpenAI SDK

Because the endpoint is OpenAI-compatible, the same SDK call reaches non-OpenAI models too. Set the model field to a Claude or Gemini model ID and the request routes through Router One — no separate client per provider.

from openai import OpenAI

# Same OpenAI SDK, calling a Claude model
client = OpenAI(base_url="https://api.router.one/v1", api_key="sk-...")

resp = client.chat.completions.create(
    model="anthropic/claude-sonnet-5",
    messages=[{"role": "user", "content": "Hello from the OpenAI SDK"}],
)
print(resp.choices[0].message.content)

Works with the tools you already use

Any client that accepts a custom OpenAI-compatible base URL works with Router One. Set the base URL and key, and keep your existing workflow.

Routing, fallback, and traces come for free

The same OpenAI-compatible endpoint gives you smart routing and resilience without extra integration work.

• Production traffic uses model_name by default, so a request is served by the model you name. Set model to "auto" and the gateway picks a low, medium or high tier for the request and serves it from that tier's server-owned model list.
• On a retryable upstream error (429, 5xx, timeout), a named-model request may try the next healthy provider serving that same model; routes with a sustained upstream-error rate are moved to the back of the order until they recover.
• Customer Logs shows the model, request ID, tokens, settled cost, total time (plus time to first token on streamed requests that produced output), status, and any sanitized error for recorded calls. Upstream provider identities, routes, and internal attempts are not disclosed.

FAQ