Integrate once. Route everywhere.

Point your OpenAI-compatible client at LLM Gateway and keep routing, provider credentials, limits, and billing policy out of the app layer.

Get API key Launch onboarding

Quickstart

API shape /v1

Drop in with OpenAI-compatible clients and keep your integration shape familiar.

Routing mode auto

Let the gateway pick provider, fallback chain, and budgets for you.

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_VIRTUAL_KEY",
    base_url="http://127.0.0.1:8000/v1",
)

response = client.chat.completions.create(
    model="auto",
    messages=[{"role": "user", "content": "Say hello"}],
)

What developers get

The goal is not just one endpoint. It is a cleaner operational layer around real LLM traffic.

01 Compatibility

OpenAI-compatible `/v1` routes plus Anthropic-style messaging support for existing clients.

02 Control

Virtual API keys, budgets, model allow-lists, and organization-scoped provider key management.

03 Visibility

Usage ledger, request logs, latency, provider attribution, and dashboard-level health visibility.

Developer flow

The fastest path from signup to a working request.

1. Create key

2. Swap base URL

Point your SDK to `/v1` and keep your client-side integration shape mostly unchanged.

3. Observe usage

Inspect provider, latency, cache status, and spend in the dashboard instead of guessing from logs.