AI Model Capability Comparison

Computer use

Only models released within the past year are shown. Only benchmarks covering at least 3 of these models are shown. Unknown release dates are excluded; historical data is retained.

Coarena · mixed tasks

— means no published score for this candidate in this source, not zero.7/11 recent-year models have scores

No.	Model	Agent engine	Reasoning effort		95% interval
1	GPT-6 Solgpt-6-sol	Coarena	medium	1043	985–1101
2	GPT-6.1 Solgpt-6.1-sol	Coarena	medium	1025	965–1085
3	GPT-6 Lunagpt-6-luna	Coarena	medium	1018	966–1070
4	Claude Opus 5.5claude-opus-5-5	Coarena	medium	1010	973–1047
5	Claude Fable 5.1claude-fable-5-1	Coarena	medium	1001	968–1034
6	Claude Sonnet 5.5claude-sonnet-5-5	Coarena	medium	989	928–1050
7	GPT-6 Astragpt-6-astra	Coarena	medium	976	936–1016

Coarena · mixed tasks

Source: Coarena ↗ · Fetched Oct 21, 2026, 06:00 UTC

Coarena · mixed tasks · Source date unknown

PandaNpc Updated Oct 21, 2026, 06:01 UTC

Source: Epoch AI, CC BY 4.0 · Vals.ai · Artificial Analysis (artificialanalysis.ai)

Column guide

No.: Display order only. Sorting changes this number; it is not a published capability rank.
Model: The evaluated model name and its provider icon. A dash in a score column means no published score is available here, not a zero.
Agent engine: Coarena
Reasoning effort: medium
Score: Blind human preference rating on mixed browser, desktop, coding and question-answering tasks in Coarena’s harness. This is a Bradley–Terry rating, not a percentage or verified task success. The interval is 95%; overlapping intervals do not establish a winner. These are overall results, not the desktop subset. Reasoning effort is shown only when published. Ranking uses the latest 20,000 eligible judged battles; PandaNpc shows its fetch time because this snapshot provides no computation timestamp. Evaluation methodology ↗
95% interval: 95% interval