Tag · 41 charts · newest first
AI Benchmarks
Our analysis and charts on AI benchmarks: coding, maths, reasoning, agents, vision and how models compare with people. Every benchmark chart, grouped by skill, is on the AI Benchmarks hub.
-
AI BenchmarksClaude Sonnet 4.6 led Finance Agent at 63.3% in early 2026; no model reached two-thirds19 Oct 2026 · Chart · CSVView →
-
AI BenchmarksGPT 5.1 led CaseLaw v2 legal research at 73.4% in early 2026, 3.5 points ahead of GPT 4.119 Oct 2026 · Chart · CSVView →
-
AI BenchmarksAll top 15 models scored above 83% on LegalBench in early 2026; Gemini 3.1 Pro Preview led at 87.4%19 Oct 2026 · Chart · CSVView →
-
AI BenchmarksThe best model scored 74.5% on GAIA in September 2025, 17.5 points short of the 92% human baseline19 Oct 2026 · Chart · CSVView →
-
AI BenchmarksThe best AI agent completed 66.3% of OSWorld computer tasks in 2025, against 72.4% for people19 Oct 2026 · Chart · CSVView →
-
AI BenchmarksWeb agents completed 74.3% of WebArena tasks in early 2026, within 4 points of people's 78.2%19 Oct 2026 · Chart · CSVView →
-
AI BenchmarksThe best agent succeeded on 64.4% of MLE-bench's Kaggle tasks in early 2026, up from 17.1% in 202419 Oct 2026 · Chart · CSVView →
-
AI BenchmarksAI agents solved 93% of Cybench security challenges unguided in early 2026, up from 17.5% in 202419 Oct 2026 · Chart · CSVView →
-
AI BenchmarksClaude Opus 4.5 led τ-bench's tool-use tasks at 70.2% in early 2026; the top seven span 7.3 points19 Oct 2026 · Chart · CSVView →
-
AI BenchmarksThe best SWE-bench Verified score reached 98.7% of the human baseline in 2025, up from 64.3%19 Oct 2026 · Chart · CSVView →
-
AI BenchmarksGemini-3.1-Pro leads MMLU-Pro at 91.2%, and the top 15 models are within 4.2 points19 Oct 2026 · Chart · CSVView →
-
AI BenchmarksClaude Opus 4.6 tops the Text Arena at 1,503, and the top 15 models span 46 points19 Oct 2026 · Chart · CSVView →