Benchmarks
Component benchmark scores with methodology notes and per-source links, plus intelligence-versus-price and intelligence-versus-speed scatters.
No data yet
This page renders from sources that have not been ingested. It is empty because the data is absent, not because the value is zero.
- ·Artificial Analysis component scores — GPQA, HLE, LiveCodeBench, SciCode, AIME
- ·LMArena Elo with confidence intervals
- ·Epoch benchmarking hub series
Blocked on: the three LOGIN roles in .env.example. Until they exist no adapter can run. Artificial Analysis also needs its API key, and is internal-plane only under v3 §6.1.