Build on Simon with one OpenAI-compatible API

Choose how you build

Two paths on the same API

Model lineup

Experts, orchestrators, and direct access

Simon routes to the right model automatically — or pin a slug and we upgrade the model behind it without migrations.

T1–T4 tiers map to simon-says-* orchestrators. Direct provider slugs (claude-opus-4-8, gpt-5.5, etc.) bill at upstream pass-through rates listed in Platform and on /models.

Full model catalog & availability →Compare vs direct provider pricing →

Build by task

Start with what you're shipping

Reference

Your first request

Works with the OpenAI SDK — set baseURL to api.geniuspro.io/v1 and use your GeniusPro API key from Platform.

curl https://api.geniuspro.io/v1/chat/completions \
  -H "Authorization: Bearer $GENIUSPRO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "simon-says-chat",
    "messages": [{"role": "user", "content": "Hello, Simon."}]
  }'

curl https://api.geniuspro.io/v1/chat/completions \
  -H "Authorization: Bearer $GENIUSPRO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "cat:my-support-bot",
    "messages": [{"role": "user", "content": "Reset my password."}]
  }'

Python, Node, and any OpenAI-compatible client — point baseURL at api.geniuspro.io/v1. Full SDK guides live in the docs portal.

Use cases

What teams build on Simon

Enterprise-ready

Built for production workloads

No training on your data, PII scrubbed in logs, and US-hosted upstream providers — with full request telemetry in Platform.

Security & compliance →Enterprise & white label →

Related products

Ship faster with the full stack

Simon is the API — these products extend it for teams, workflows, and infrastructure.