Benchmark · agentic

CursorBench

3 results 3 models

Cursor's private, version-tagged internal agent eval built from real Cursor sessions; the trailing number is a version, not a score. Not one of the public suites it is quoted beside.

0 19.5 39 58.5 78 2026-07-31 2026-08-16 2026-09-01 Claude Opus 5 · 70 · 2026-07-31 Grok 4.6 · 69.9 · 2026-08-13 Claude Fable 5.1 · 73.4 · 2026-09-01
Claude Opus 5 Grok 4.6 Claude Fable 5.1
Timeline
Date Model Score Source
2026-09-01 Claude Fable 5.1 73.4% Anthropic releases Claude Fable 5.1 with 52.6% on Terminal-Bench-Science and cheaper cache reads
2026-08-13 Grok 4.6 69.9% SpaceXAI releases Grok 4.6 with 500K context and improved agentic reasoning
2026-07-31 Claude Opus 5 70.0% Anthropic launches Claude Opus 5; OpenAI models breach Hugging Face during security test