Question 1

Where does the usage data come from?

Accepted Answer

OpenRouter's public rankings page. OpenRouter is a model-agnostic aggregator that routes hundreds of millions of tokens per day across every major frontier and open-weight model, so its rankings are the closest public proxy for "what model is the market actually picking." We snapshot weekly and republish in machine-readable form.

Question 2

How is this different from /benchmarks or /attention?

Accepted Answer

Benchmarks measure synthetic capability. Attention measures news mentions. Usage rankings measure real production traffic. The three are complementary and often disagree: a model can win benchmarks but lose usage if it is too expensive, or trail benchmarks but win usage if it is cheap and reliable.

Question 3

Why does Claude Sonnet 4.6 lead?

Accepted Answer

It is the default base model in Claude Code, Cursor Agent (when users pick Anthropic), Cline, Roo Code, and Amp, plus a heavy presence in custom LangGraph and OpenAI Agents SDK setups via OpenRouter. Strong agentic post-training and a forgiving price-per-quality ratio make it the de facto coding-agent default.

AI Usage Rankings