- input
- $0.20
- cached
- $0.03
- output
- $1.10
- Context
- 131Ktokens
- Released
- 1y ago
- Size
- 110Bparameters
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 #201
- 1453
GPT-5.6 LunaSimilar price - 1417
DeepSeek V3.1Similar price - 1375
Gemini 2.5 Flash-Lite Preview 06-17Nearby rank - 1374
Claude Sonnet 3.5 v2Nearby rank - 1373
GLM 4.5 Air - 1369
Qwen3-Next 80B-A3B (Thinking)Nearby rank - 1365Gemma 3 27BNearby rank
- 1346
Qwen-PlusSimilar price - 1303
Mistral Small 3.1 24BSimilar price
Ahead of 2 of 4 similarly priced models
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
Humanity's Last Exam
Best score per model · Higher is better
#24 of 34
- 19.4%GPT-5 MiniSimilar price
- 8.6%Gemini 3.1 Flash LiteSimilar price
- 8.3%GLM 4.5Nearby rank
- 8.1%GLM 4.5 Air
- 8.0%Claude Sonnet 3.7Nearby rank
- 7.9%Claude Opus 4.1Nearby rank
- 5.7%
Llama 4 MaverickSimilar price - 4.5%Mistral Medium 3Similar price
Ahead of 2 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 #2011373
- default137395% CI 1369 to 1377 · 30,591 votes
- Humanity's Last Exam#24 of 348.1%
- default8.1%
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.