SWE-bench Verified
500 human-validated GitHub issue resolution tasks from SWE-bench, created with OpenAI in August 2024. The standard evaluation for agentic coding systems. Top performers (2026) exceed 78% resolved.
Models ranked
104
tracked on this benchmark
Score band
96.0 – 23.6
best → lowest tracked
Snapshot trend
-10.25
Oct 21 → Nov 6 · 2 models
Leaderboard
Tracked models ranked by % Resolved (higher is better).
Notes: Independent Vals.ai SWE-bench Verified result; keep separate from Anthropic's SWE-bench Pro launch row.
Configuration: Adaptive thinking at maximum effort
Notes: Vendor-reported. Source: p.149, Table 8.1.A. Preserve the five-trial setting; do not treat single-trial rows as equivalent.
Notes: Official Anthropic launch benchmark.
Configuration: Ornith-1.5-122B
Notes: Self-reported by Ornith. Ornith first-party HF model card Ornith-1.5-122B, benchmark table (column 'Ornith-1.5-122B'). Retrieved 2026-11-06. SWE-bench Verified (non-hack): 84.5.
Notes: source reconciliation: vals.ai standardized harness reports GPT-5.5 at 82.6; do not use the non-comparable 88.7 self-reported row for /best/coding.
Notes: Medium confidence: GPT-5.5 Pro uses the same underlying weights as standard GPT-5.5; no Pro-specific SWE-bench Verified score was found.
Notes: Reflection first-party final model card (HF https://huggingface.co/reflection/Beam-1.0-501B-A23B, BF16 column), retrieved 2026-10-16. Replaces 2026-10-05 preview-blog score 80.9. Card label 'SWEBench Verified'. Lab self-reported.
Notes: Confidence: medium. May 12 /best/ refresh; benchlm.ai May 11 snapshot showed Claude Opus 4.5 at 80.9%.
Notes: Verified via third-party source; official OpenAI benchmark may differ. Use directionally.
Notes: Confidence: high. Anthropic self-reported this score and identified Composer 2.5's prior 80.8 SWE-bench Verified row as a Claude Opus 4.6 misattribution.
Notes: Confidence: high. May 12 /best/ refresh; benchlm.ai and llm-stats both reported Qwen3.6 Plus at 78.8%.
How to read this benchmark
This benchmark scores models where higher is better. Use scores for directional filtering and shortlisting, not universal quality ranking; then validate pricing, context window, provider availability, and fit for your workload.
Trust this score when
- There is a fresh timestamped snapshot (or multiple snapshots) for this benchmark.
- The model list covers the same version family you can actually deploy today.
- Top candidates overlap with your required routing and feature requirements.
Be cautious when
- There is only one benchmark snapshot or the dataset appears stale.
- The benchmark metric direction is opposite of your decision objective.
- The score difference between options is narrow and likely within implementation variance.
Related benchmarks
Last reviewed: Apr 15, 2026