- input
- $0.03
- cached
- $0.03
- output
- $0.13
- Context
- 131Ktokens
- Released
- 1y ago
- Size
- 20Bparameters
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 #271
- 1474
GLM 5.3 FlashSimilar price - 1318
Llama 3.3 70B InstructNearby rank - 1318
GPT-4o-mini (2024-07-18)Nearby rank - 1318gpt-oss-20b
- 1317
Gemma 3n E4B ITNearby rank - 1314
Mistral Large 2407Nearby rank - 1303Gemma 3 4BSimilar price
- 1274Mistral Small 3Similar price
- 1241
Nova Micro 1.0Similar 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
GPQA diamond
Best score per model · Higher is better
#121 of 161
- 90.2%GLM 5.3 FlashSimilar price
- 82.3%
Qwen3.7 FlashSimilar price - 63.8%Qwen3 14BNearby rank
- 61.7%Qwen3 30B A3BNearby rank
- 60.8%gpt-oss-20b
- 60.5%GLM 4.7 FlashNearby rank
- 59.5%Mistral Medium 3Nearby rank
- 23.9%Llama 3.2 1B InstructSimilar price
- 23.2%Gemma 3 4BSimilar price
Ahead of 2 of 4 similarly priced models
OTIS Mock AIME 2024-2025
Best score per model · Higher is better
#108 of 158
- 93.9%GLM 5.3 FlashSimilar price
- 86.7%Qwen3.7 FlashSimilar price
- 66.4%
R1 0528Nearby rank - 66.4%Qwen3 14BNearby rank
- 65.3%gpt-oss-20b
- 64.4%
Claude Opus 4Nearby rank - 62.8%Qwen3 30B A3BNearby rank
- 7.5%Gemma 3 4BSimilar price
- 0.6%Llama 3.2 1B InstructSimilar price
Ahead of 2 of 4 similarly priced models
Terminal Bench
Best score per model · Higher is better
#49 of 49
- 17.1%Gemini 2.5 FlashNearby rank
- 9.2%Qwen3.5-9BSimilar price
- 3.4%gpt-oss-20b
Ahead of 0 of 1 similarly priced model
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.
Reasoning effort
GPQA diamond
Score by reasoning effort · Higher is better
OTIS Mock AIME 2024-2025
Score by reasoning effort · Higher is better
All published scores
- Arena TextArena #2711318
- default131895% CI 1311 to 1324 · 10,400 votes
- GPQA diamond#121 of 16160.8%
- low53.2%
- medium60.8%
- high46.0%
Source:Epoch AI evaluation - OTIS Mock AIME 2024-2025#108 of 15865.3%
- low40.3%
- medium65.3%
- high50.8%
Source:Epoch AI evaluation - Terminal Bench#49 of 493.4%
- default3.4%Mini-SWE-AgentTerminal-Bench v2 LeaderboardOriginal-source terms apply
- unknown3.4%Mini-SWE-Agenthttps://www.tbench.ai/leaderboard/terminal-bench/2.0Original-source terms apply
- default3.1%Terminus 2Terminal-Bench v2 LeaderboardOriginal-source terms apply
- unknown3.1%Terminus 2https://www.tbench.ai/leaderboard/terminal-bench/2.0Original-source terms apply
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.