- input
- $0.40
- cached
- $0.04
- output
- $2.00
- Context
- 131Ktokens
- Released
- 1y ago
Cost/1M tokens
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 #188
- 1455
Gemini 3.5 Flash LiteSimilar price - 1434
MiMo-V2.5Similar price - 1409
Qwen3.5-27BSimilar price - 1388
Claude Sonnet 3.7Nearby rank - 1387
Solar Pro 4Nearby rank - 1387
Mistral Medium 3 - 1383
GPT-4.1 MiniNearby rank - 1382Qwen3 30B A3B Instruct 2507Nearby rank
- 1335
Nova 2 LiteSimilar price
Ahead of 1 of 4 similarly priced models
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
Arena #116
- 1266Gemini 3.5 Flash LiteSimilar price
- 1234MiMo-V2.5Similar price
- 1216Qwen3.5-27BSimilar price
- 1160Gemma 3 27BNearby rank
- 1160Mistral Medium 3.1Nearby rank
- 1155Mistral Medium 3
- 1153
GLM 4.5VSimilar price - 1147GPT-5 NanoNearby rank
Ahead of 0 of 4 similarly priced models
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-2025-02-28-Private
Best score per model · Higher is better
#52 of 58
- 26.2%Qwen3.6 PlusSimilar price
- 4.8%Gemini 2.5 FlashSimilar price
- 4.5%GPT-4.1 MiniSimilar price
- 3.8%GLM 4.6Similar price
- 1.0%
Qwen MaxNearby rank - 1.0%GPT-4.1 NanoNearby rank
- 0.3%Mistral Medium 3
- 0.3%Claude Haiku 3.5Nearby rank
- 0.3%Mistral Large 2.1Nearby rank
Ahead of 0 of 4 similarly priced models
GPQA diamond
Best score per model · Higher is better
#123 of 161
- 88.4%Qwen3.6 PlusSimilar price
- 83.3%Gemini 3.5 Flash LiteSimilar price
- 76.3%
R1 0528Similar price - 60.8%gpt-oss-20bNearby rank
- 60.5%GLM 4.7 FlashNearby rank
- 59.5%Mistral Medium 3
- 56.8%Qwen3 8BNearby rank
- 56.1%Qwen MaxNearby rank
- 44.2%
Llama 3.1 70B InstructSimilar price - 15.2%Mistral 7B Instruct v0.3Predecessor
+44.4 pts vs Mistral 7B Instruct v0.3 (predecessor) · Ahead of 1 of 4 similarly priced models
Humanity's Last Exam
Best score per model · Higher is better
#32 of 34
- 24.4%
Kimi K2.5Similar price - 19.4%GPT-5 MiniSimilar price
- 8.6%Gemini 3.1 Flash LiteSimilar price
- 8.3%GLM 4.5Similar price
- 5.5%Claude Sonnet 4Nearby rank
- 5.4%GPT-4.1Nearby rank
- 4.5%Mistral Medium 3
- 4.1%Claude Sonnet 3.5 v2Nearby rank
- 2.7%GPT-4o (2024-11-20)Nearby rank
Ahead of 0 of 4 similarly priced models
OTIS Mock AIME 2024-2025
Best score per model · Higher is better
#126 of 158
- 93.3%Qwen3.6 PlusSimilar price
- 71.1%Gemini 3.5 Flash LiteSimilar price
- 66.4%R1 0528Similar price
- 38.3%GPT-4.1Nearby rank
- 37.8%DeepSeek V3 0324Nearby rank
- 32.2%Mistral Medium 3
- 30.0%Magistral SmallNearby rank
- 28.9%GPT-4.1 NanoNearby rank
- 3.6%Llama 3.1 70B InstructSimilar price
- 0.3%Mistral 7B Instruct v0.3Predecessor
+31.9 pts vs Mistral 7B Instruct v0.3 (predecessor) · Ahead of 1 of 4 similarly priced models
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 #1881387
- default138795% CI 1383 to 1392 · 32,751 votes
- Arena VisionArena #1161155
- default115595% CI 1147 to 1163 · 10,773 votes
- FrontierMath-2025-02-28-Private#52 of 580.3%
- default0.3%
Source:Epoch AI evaluation - GPQA diamond#123 of 16159.5%
+44.4 pts vs Mistral 7B Instruct v0.3 (predecessor)
- default59.5%
Source:Epoch AI evaluation - Humanity's Last Exam#32 of 344.5%
- default4.5%
- OTIS Mock AIME 2024-2025#126 of 15832.2%
+31.9 pts vs Mistral 7B Instruct v0.3 (predecessor)
- default32.2%
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.
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.