docs(benchmarks): add Terminal Bench 2.1 results page - #14

Closed

andreyrisukhin-factory wants to merge 13 commits into

mainfrom

ay/docs-terminal-bench-2-1

Commits on Jul 25, 2026

docs: correct Claude Opus 5 multiplier to 2x (#1333)

factory-austin
and
factory-droid[bot]
authored
33926e5
docs: add CLI changelog entry for v0.180.0 (#1332)
factory-droid[bot]
authored
7ea5f9c

Commits on Jul 27, 2026

docs(cli): distinguish /clear from /new in CLI reference (#1335)

factory-arman
and
factory-droid[bot]
authored
18366ae
docs: add Kimi K3 to available models (#1317)

authored
0979803

Commits on Aug 17, 2026

docs(benchmarks): add Terminal Bench 2.1 results page

authored and
Andrey Risukhin
committed
61d73b0
docs(benchmarks): add accuracy-vs-cost frontier plot to TB2.1 page

authored and
Andrey Risukhin
committed
747c0f4
docs(benchmarks): use per-session billed ledger for TB2.1 cost plot

authored and
Andrey Risukhin
committed
dbadb67
docs(benchmarks): add Droid reasoning effort to TB2.1 table

authored and
Andrey Risukhin
committed
c3fdade
docs(benchmarks): add per-model mean cost-accuracy view to TB2.1

authored and
Andrey Risukhin
committed
cd70799
docs(benchmarks): exclude kimi's stale 07-29 cohort from TB2.1

authored and
Andrey Risukhin
committed
bedea82
docs(benchmarks): fix $ cost rendering and soften single-winner framing

authored and
Andrey Risukhin
committed
3a4c25b
docs(benchmarks): add 95% CIs to mean view, strengthen trend framing

authored and
Andrey Risukhin
committed
59443b6
docs(benchmarks): switch mean-plot bars to run range, fold in matched-effort runs

authored and
Andrey Risukhin
committed
0861c91