Opus Brain,
GPT Speed
One API call. Opus thinks and reviews, GPT codes and executes. 20% cheaper than official rates, same quality. Just swap the endpoint.
A Smarter Backend for OpenClaw
OpenClaw needs an LLM backend to power its agent workflows. With AI Duet, Opus handles decisions and review, GPT handles coding and execution. Every call gets Opus-level quality, 20% cheaper.
Swap the Endpoint
Replace the API URL in your OpenClaw config with api.aiduet.app and enter your AI Duet key.
Auto Orchestration
Every instruction is analyzed by Opus, with coding tasks automatically delegated to GPT for high-speed execution.
Quality + Savings
Opus reviews GPT's output to ensure quality. Overall token usage is 60% less than pure Opus.
api_url = "https://api.aiduet.app"
api_key = "hr-your-key-here"
model = "opus-4.6"Why AI Duet
Opus Thinks, GPT Codes
Opus analyzes intent, breaks down tasks, reviews results. GPT takes over coding at full speed. Each model does what it's best at.
20% Cheaper
Every token is 20% below official API rates. Opus at $4/M input, GPT at $2/M input. Pay per token, no subscriptions.
Dual Format
Anthropic Messages API and OpenAI Chat API both work. OpenClaw, Cursor, your own system — just swap the endpoint.
Real-time Streaming
SSE streaming delivers Opus thinking and GPT code output in real time. Same experience as a direct API connection.
Transparent Costs
Per-request token usage and cost breakdown, visible in real time. Pre-load balance, no commitments, no subscriptions.
One-line Setup
Swap the API endpoint and key. Works with OpenClaw, Cursor, Continue, Cline, and all major AI coding tools.
One Call, Three-Step Duet
Opus Plans
Opus receives the request, understands intent, and generates execute_code tool call instructions.
GPT Codes
GPT executes coding tasks at high speed, generates code and returns results. Fast and affordable.
Opus Reviews
Opus reviews code quality, confirms accuracy, then returns the final response. Loops again if needed.
Pay As You Go
No monthly fees. No hidden costs. Pre-load balance, pay per token.
Official API
Most intelligent model for agents and coding
High-speed execution model, best value
/ 1M tokens
Single model, no orchestration
AI Duet
Most intelligent model for agents and coding
High-speed execution model, best value
/ 1M tokens
Opus reviews + GPT codes, auto orchestration
- Opus + GPT smart orchestration
- Anthropic / OpenAI dual format
- SSE real-time streaming
- Transparent usage details
- Pre-load balance pay-per-use
Swap the Endpoint, Save 20%
One config change to give OpenClaw, Cursor, and more the power of Opus + GPT orchestration.
Sign Up FreeFree to sign up · Pay as you go · Cancel anytime