evals.report
BenchmarksLabsCompareRun guidesIn the wild
Microsoft AIMAI (Microsoft AI)

MAI-Thinking-1

Microsoft AI · MAI (Microsoft AI). Released Jun 2, 2026.

MAI-Thinking-1 is a model from Microsoft AI in the MAI (Microsoft AI) family, released Jun 2, 2026. evals.report tracks 5 reported MAI-Thinking-1 benchmark scores across SWE-bench Verified, SWE-bench Pro, GPQA Diamond, SimpleQA Verified, Terminal-Bench 2.0 — each shown with its benchmark, metric, source status, and date, and never combined into a single ranking.

5 results

Benchmark results 5

Compare this model
BenchmarkCategoryScoreMetricStatusDate
SWE-bench VerifiedCoding73.5%% resolvedVerifiedJun 2, 2026Details
SWE-bench ProCoding52.8%% resolvedVerifiedJun 2, 2026Details
GPQA DiamondReasoning84.2%accuracyVerifiedJun 2, 2026Details
SimpleQA VerifiedOther31%accuracyVerifiedJun 2, 2026Details
Terminal-Bench 2.0Agents46.0%task successVerifiedJun 2, 2026Details