Benchmarks

FrontierBench

Agentsunit % · normalized over [0, 50]

Frontier agent-work benchmark from the Terminal-Bench/Harbor team (formerly Terminal-Bench 3.0): 74 tasks across 7 domains in v0.1, spanning artifacts from databases and ML checkpoints to CAD files and formal proofs; resolution rate per agent-harness x model pair, with per-trial token and cost accounting.

#ModelSourceScoreNormalized

Score distribution

8 tracked results across the normalization window

050

Score vs. parameters

Open-weights models, log-x params

Only one open-weights model with a disclosed parameter count — not enough to plot a trend.