- Context
- 1Mtokens
- Released
- 8mo 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
Not ranked on Arena Text yet
Arena WebDev
Best Arena rating per model · Higher is better
Not ranked on Arena WebDev yet
Arena Vision
Best Arena rating per model · Higher is better
Not ranked on Arena Vision yet
Arena Search
Best Arena rating per model · Higher is better
Not ranked on Arena Search yet
Arena Document
Best Arena rating per model · Higher is better
Not ranked on Arena Document yet
Agent Arena
Best IPS score per model · Higher is better
Not ranked on Agent Arena yet
FrontierMath-Tier-4-v2-Private
Best score per model · Higher is better
#51 of 72
- 58.5%
GPT-5.4 ProSimilar release date - 26.8%
Claude Opus 4.6Similar release date - 26.8%
Gemini 3.1 Pro PreviewSimilar release date - 25.6%
Kimi K2.6Similar release date - 17.1%
Inkling SmallNearby rank - 17.1%
Grok 4.7Nearby rank - 17.1%Grok 4.20 (Reasoning)
- 17.1%
GLM 5.3 FlashNearby rank - 14.6%Grok 4.3Nearby rank
Ahead of 0 of 4 models released around the same time
FrontierMath-Tiers-1-3-v2-Private
Best score per model · Higher is better
#58 of 87
- 82.5%GPT-5.4 ProSimilar release date
- 59.6%Gemini 3.1 Pro PreviewSimilar release date
- 45.3%
DeepSeek V4 Pro 0423Nearby rank - 44.9%GPT-5.4 NanoNearby rank
- 44.9%Grok 4.20 (Reasoning)
- 42.8%Grok 4.3Nearby rank
- 38.2%
Qwen3.6 PlusNearby rank - 36.8%GLM 5.1Similar release date
- 18.2%Qwen3.5-FlashSimilar release date
Ahead of 2 of 4 models released around the same time
GPQA diamond
Best score per model · Higher is better
#49 of 161
- 94.6%GPT-5.4 ProSimilar release date
- 94.4%Gemini 3.1 Pro PreviewSimilar release date
- 89.6%Claude Haiku 5.5Nearby rank
- 89.4%Gemini 3 Flash PreviewNearby rank
- 89.3%Grok 4.20 (Reasoning)
- 88.8%Grok 4.3Nearby rank
- 88.5%Inkling SmallNearby rank
- 87.4%Claude Sonnet 4.6Similar release date
- 82.3%Qwen3.5-FlashSimilar release date
Ahead of 2 of 4 models released around the same time
OTIS Mock AIME 2024-2025
Best score per model · Higher is better
#51 of 158
- 97.8%GPT-5.4Similar release date
- 95.6%Gemini 3.1 Pro PreviewSimilar release date
- 93.3%Grok 4.3Nearby rank
- 93.3%GLM 5.1Nearby rank
- 92.2%Grok 4.20 (Reasoning)
- 91.4%Gemini 3 Pro PreviewNearby rank
- 91.4%GPT-5Nearby rank
- 84.4%Qwen3.5-FlashSimilar release date
- 55.8%
Qwen3.5 4BSimilar release date
Ahead of 2 of 4 models released around the same time
SimpleQA Verified
Best score per model · Higher is better
#68 of 87
- 73.5%Gemini 3.1 Pro PreviewSimilar release date
- 46.3%GPT-5.4 ProSimilar release date
- 35.5%Claude Sonnet 4.6Similar release date
- 31.1%GPT-4.1Nearby rank
- 30.7%Claude Sonnet 4.5Nearby rank
- 30.2%Grok 4.20 (Reasoning)
- 29.4%GPT-5.4 MiniNearby rank
- 26.0%GPT-4o (2024-08-06)Nearby rank
- 20.3%Qwen3.5-FlashSimilar release date
Ahead of 1 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
- FrontierMath-Tier-4-v2-Private#51 of 7217.1%
- default17.1%
Source:Epoch AI evaluation - FrontierMath-Tiers-1-3-v2-Private#58 of 8744.9%
- default44.9%
Source:Epoch AI evaluation - GPQA diamond#49 of 16189.3%
- default89.3%
Source:Epoch AI evaluation - OTIS Mock AIME 2024-2025#51 of 15892.2%
- default92.2%
Source:Epoch AI evaluation - SimpleQA Verified#68 of 8730.2%
- default30.2%
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.