BlogLLM Benchmark Scores as Published, Side by Side
CursorBench: Results as Published
Every result the catalog holds for CursorBench: 10 models from 3 developers, read from 9 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Agentic CodingMaintainer Cursor (Anysphere, Inc.)Introduced 2026Data version 2026103001, as of 2026-10-30
CursorBench is Cursor's own coding-agent benchmark, built from ambiguous, multi-file tasks taken from real Cursor sessions. Version 4.0 adds long-horizon work such as editing, refactoring, investigation and managing jobs. Each agent gets a percentage score, and the page also reports cost per task and tokens used.
CursorBench 4.0
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-5.6 SolOpenAI, released 2026-07-09 | 41.7% | Anthropic, Claude Opus 5.5checked 2026-10-14; printed on 2 more pages | - |
| Claude Sonnet 5.5Anthropic, released 2026-09-28 | 55.5% | Anthropic, Claude Sonnet 5.5checked 2026-10-14; printed on 1 more page | - |
| Claude Opus 5.5Anthropic, released 2026-09-22 | 52.5% | Anthropic, Claude Opus 5.5checked 2026-10-25; printed on 2 more pages | Effort: medium |
| 57.8% | Anthropic, Claude Opus 5.5checked 2026-10-14 | - | |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 51.8% | Anthropic, Claude Opus 5.5checked 2026-10-14; printed on 2 more pages | - |
| Claude Opus 5Anthropic, released 2026-07-24 | 46.6% | Anthropic, Claude Opus 5.5checked 2026-10-14; printed on 1 more page | - |
| Claude Sonnet 5Anthropic, released 2026-06-30 | 34.1% | Anthropic, Claude Sonnet 5.5checked 2026-10-14 | - |
| Grok 4.7xAI, released 2026-09-21 | 41.6% | xAI, Grok 4.7checked 2026-10-25; printed on 1 more page | Effort: medium |
| 33.1% | xAI, Grok 4.7checked 2026-10-25 | Effort: low | |
| 43.9% | xAI, Grok 4.7checked 2026-10-25 | Effort: high | |
| 46.3% | xAI, Grok 4.7checked 2026-10-25 | Effort: xhigh | |
| Grok 4.6xAI, released 2026-08-12 | 40.4% | xAI, Grok 4.7checked 2026-10-25 | Effort: high |
CursorBench 3.2.0
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-5.6 SolOpenAI, released 2026-07-09 | 67.2% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 2 more pages | - |
| Claude Fable 5.1Anthropic, released 2026-09-01 | 73.4% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 1 more page | - |
| Claude Opus 5Anthropic, released 2026-07-24 | 70% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 1 more page | - |
| Claude Fable 5Anthropic, released 2026-06-09 | 70.5% | Anthropic, Claude Fable 5.1 and Mythos 5.1checked 2026-10-15; printed on 2 more pages | - |
| Grok 4.6xAI, released 2026-08-12 | 69.9% | xAI, Introducing Grok 4.6checked 2026-10-15 | Effort: high |
| Grok 4.5xAI, released 2026-07-16 | 66.7% | xAI, Introducing Grok 4.6checked 2026-10-15 | Effort: high |
CursorBench 4.0 average cost per task
lower is better
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| Grok 4.7xAI, released 2026-09-21 | $3.49 | xAI, Grok 4.7checked 2026-10-25 | Effort: medium |
| $1.58 | xAI, Grok 4.7checked 2026-10-25 | Effort: low | |
| $4.69 | xAI, Grok 4.7checked 2026-10-25 | Effort: high | |
| $6.01 | xAI, Grok 4.7checked 2026-10-25 | Effort: xhigh |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.