Model Brief

Benchmark history

Every reading we have recorded, one chart per index.

One reading so far, so each chart shows where the models sit rather than a trend. A line starts with the second reading, and it breaks where the index version changes, because a new version is a different measurement.

Intelligence · 118 models · one reading
2 18 33 49 65 09 Nov
Coding · no reading
No reading for this index
Agentic · no reading
No reading for this index
Finance and accounting · 130 models · one reading
4 19 35 50 65 09 Nov
Strategy and operations · 134 models · one reading
3 19 35 51 67 09 Nov
Legal · 130 models · one reading
5 19 34 49 63 09 Nov
Healthcare and medical · 99 models · one reading
6 21 36 50 65 09 Nov
Engineering · 118 models · one reading
6 23 39 56 73 09 Nov
Economics · 130 models · one reading
5 20 36 51 66 09 Nov

Read 9 November 2026. Data: Artificial Analysis (artificialanalysis.ai). The free tier publishes the intelligence index and seven domain indices; coding and agentic come back empty for every model, so those two charts stay empty.