From the IQBANDIT marketplace to your first live query in under a minute. Pick a model, pick an integration, and you're live — no GPU bills, no infrastructure headaches.
Which model do you want to use?
Which integration do you want to setup?
Sign in to start your free trial.
Infrastructure
Your OpenClaw gateway runs on your own infrastructure. You own the tokens, the traffic, and the logs — we never touch your data.
# Start your OpenClaw gateway
$ openclaw start --port 19001
✓ Gateway listening on http://127.0.0.1:19001
✓ Auth token ready
✓ Health check passed
$ curl http://127.0.0.1:19001/v1/chat/completions \
-H "Authorization: Bearer $TOKEN"
-d '{"model":"claude-opus-4"}'
< 200 OK — response streaming
Q3 Summary
Build & iterate
Every agent runs on its own endpoint. Query it from the browser playground, call it from your codebase, or connect any OpenAI-compatible client directly to your gateway.
Pre-built agents handle real workflows out of the box.
Choose from Claude, GPT-4o, Gemini, or Mistral. Select Playground, API, Discord, or Telegram.
Sign in, connect your OpenClaw gateway, and your agent is reachable in under 60 seconds.
Your agent is deployed and ready. Query it via the playground or start building on the API.