- Context
- 262Ktokens
- Released
- 1y ago
Benchmarks
DeepSWE is the primary coding benchmark where an exact model evaluation exists. Arena leaderboards and other sourced evaluations appear below; missing scores are not zero.
How it ranks
Against its predecessor, similarly priced models and the models ranked around it.
DeepSWE 1.1
Best pass@1 per model · Higher is better
Not evaluated in DeepSWE 1.1
Arena Text
Best Arena rating per model · Higher is better
Arena #128
- 1456
Claude Sonnet 4.5Similar release date - 1425
DeepSeek V3.2Nearby rank - 1424
GLM 4.6Nearby rank - 1423
Qwen3 Max - 1423
Mistral Large 4Nearby rank - 1422Claude Haiku 5.5Nearby rank
- 1416DeepSeek V3.1 TerminusSimilar release date
- 1413
Qwen3 VL 235B A22B InstructSimilar release date - 1379
Gemini 2.5 Flash-Lite Preview 09-2025Similar release date
Ahead of 1 of 4 models released around the same time
Arena WebDev
Best Arena rating per model · Higher is better
Not ranked on Arena WebDev yet
Arena Search
Best Arena rating per model · Higher is better
Not ranked on Arena Search yet
Agent Arena
Best IPS score per model · Higher is better
Not ranked on Agent Arena yet
FrontierMath-Tiers-1-3-v2-Private
Best score per model · Higher is better
#80 of 87
- 55.8%
GPT-5 ProSimilar release date - 51.2%Gemini 3 Flash PreviewSimilar release date
- 31.2%Qwen3.5 397B A17BSimilar release date
- 23.9%Claude Sonnet 4.5Similar release date
- 20.0%GPT-5 NanoNearby rank
- 19.3%Qwen3.7 FlashNearby rank
- 18.9%Qwen3 Max
- 18.2%Qwen3.5-FlashNearby rank
- 12.6%Claude Opus 4.1Nearby rank
Ahead of 0 of 4 models released around the same time
GPQA diamond
Best score per model · Higher is better
#102 of 161
- 84.2%
Kimi K2 Thinking TurboSimilar release date - 82.3%Claude Sonnet 4.5Similar release date
- 77.6%
Seed 1.6Similar release date - 75.0%GPT-5 MiniNearby rank
- 73.2%Gemma 4 26B A4BNearby rank
- 72.6%Qwen3 Max
- 71.5%Seed-OSS-36B-InstructNearby rank
- 71.2%Claude Haiku 4.5Nearby rank
- 47.6%Magistral Small 1.2Similar release date
Ahead of 1 of 4 models released around the same time
OTIS Mock AIME 2024-2025
Best score per model · Higher is better
#89 of 158
- 83.1%Kimi K2 Thinking TurboSimilar release date
- 80.0%GLM 5Nearby rank
- 77.8%Claude Sonnet 4.5Similar release date
- 73.3%Qwen3 Max
- 73.3%Gemma 4 31BNearby rank
- 71.1%Claude Sonnet 4Nearby rank
- 69.4%Seed 1.6Similar release date
- 28.1%Magistral Small 1.2Similar release date
Ahead of 2 of 4 models released around the same time
SimpleQA Verified
Best score per model · Higher is better
#33 of 87
- 50.1%GPT-5Similar release date
- 49.4%o3Nearby rank
- 49.3%
Grok 4.6Nearby rank - 48.7%Qwen3 Max
- 48.3%Grok 4.5Nearby rank
- 48.0%GPT-5.1Nearby rank
- 40.4%Qwen3 235B A22B Thinking 2507Similar release date
- 30.7%Claude Sonnet 4.5Similar release date
- 17.4%Seed 1.6Similar release date
Ahead of 3 of 4 models released around the same time
Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.
All published scores
- Arena TextArena #1281423
- default142395% CI 1417 to 1430 · 9,004 votes
- FrontierMath-Tiers-1-3-v2-Private#80 of 8718.9%
- default18.9%
Source:Epoch AI evaluation - GPQA diamond#102 of 16172.6%
- default72.6%
Source:Epoch AI evaluation - OTIS Mock AIME 2024-2025#89 of 15873.3%
- default73.3%
Source:Epoch AI evaluation - SimpleQA Verified#33 of 8748.7%
- default48.7%
Source:Epoch AI evaluation
Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.
Arena ratings: Arena (lmarena-ai/leaderboard-dataset), licensed CC BY 4.0. Ratings come from pairwise human votes and are only comparable within one leaderboard; the interval is Arena's 95% confidence interval. Model names are matched to the Vertical Tabs catalog.