The edge-native
AI gateway.
Route to 21+ models from 10+ providers. OpenAI-compatible API with intelligent routing, semantic caching, and auto-fallback — deployed at the edge.
PythonOpenAI Compatible
from openai import OpenAI
client = OpenAI(
base_url="https://anyrouter.dev/api/v1",
api_key="ar-...",
)
response = client.chat.completions.create(
model="google/gemma-4-26b-a4b-it:free",
messages=[{"role": "user", "content": "Hello!"}],
)Available Models
Compare pricing, latency, and capabilities across all supported AI models
Showing 21 of 21 models
21+
AI Models
10+
Providers
2
Model Categories
100%
OpenAI Compatible
Quick start
Get started in 3 steps
Drop AnyRouter into any OpenAI-compatible client — no SDK changes required.
1
Get Your API Key
Sign up and generate an API key from the dashboard in seconds
2
Point Your SDK
Change your base URL to anyrouter.dev/api — works with any OpenAI-compatible SDK
3
Route to Any Model
Access 21+ models across 10 providers with intelligent routing and caching
.env
# Export your AnyRouter credentials
ANYROUTER_API_KEY=ar_...
ANYROUTER_BASE_URL=https://anyrouter.dev/api/v1