- input
- $0.75
- cached
- $0.075
- output
- $4.50
- Context
- 400Ktokens
- Released
- 8mo 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 #87
- 1497
Gemini 3.8 FlashSimilar price - 1475
GLM 5.2Similar price - 1448
MiMo-V2-ProNearby rank - 1448MiMo-V2.6-FlashNearby rank
- 1447
GPT-5.4 Mini - 1446
ERNIE 5.0Nearby rank - 1446Gemini 2.5 ProNearby rank
- 1443
InklingSimilar price - 1442
Grok 4.3Similar price - 1390GPT-5 MiniPredecessor
+58 vs GPT-5 Mini (predecessor) · Ahead of 0 of 4 similarly priced models
Arena WebDev
Best Arena rating per model · Higher is better
Arena #107
- 1608GLM 5.2Similar price
- 1583Gemini 3.8 FlashSimilar price
- 1414InklingSimilar price
- 1399GLM 5V TurboNearby rank
- 1398
MiniMax M2.7Nearby rank - 1398GPT-5.4 Mini
- 1397GPT-5.1Nearby rank
- 1393
Claude Sonnet 4.5Nearby rank - 1357Grok 4.3Similar price
Ahead of 1 of 4 similarly priced models
Arena Vision
Best Arena rating per model · Higher is better
Arena #54
- 1293Gemini 3.8 FlashSimilar price
- 1252
Kimi K2.5Nearby rank - 1251GPT-5.1Nearby rank
- 1250GPT-5.4 Mini
- 1249Grok 4.7Nearby rank
- 1248MiMo-V2.6-FlashNearby rank
- 1245
Step 5 PreviewSimilar price - 1242Grok 4.3Similar price
- 1232GLM 5V TurboSimilar price
- 1185GPT-5 MiniPredecessor
+65 vs GPT-5 Mini (predecessor) · Ahead of 1 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
ARC-AGI-2
Best score per model · Higher is better
#50 of 85
- 89.2%Gemini 3.8 FlashSimilar price
- 65.1%Grok 4.20Similar price
- 36.5%InklingSimilar price
- 31.1%Gemini 3 Pro PreviewNearby rank
- 22.8%GLM 5.2Similar price
- 18.9%GPT-5.4 Mini
- 18.3%GPT-5 ProNearby rank
- 17.6%GPT-5.1Nearby rank
- 4.4%GPT-5 MiniPredecessor
+14.5 pts vs GPT-5 Mini (predecessor) · Ahead of 0 of 4 similarly priced models
FrontierMath-2025-02-28-Private
Best score per model · Higher is better
#16 of 58
- 39.0%Kimi K2.6Similar price
- 32.4%Claude Sonnet 4.6Nearby rank
- 31.0%GPT-5.1Nearby rank
- 28.3%GPT-5.4 Mini
- 27.2%GPT-5 MiniPredecessor
- 26.2%
Qwen3.6 PlusNearby rank - 16.4%GLM 5Similar price
- 8.5%Qwen3 235B A22B Thinking 2507Similar price
- 5.9%Claude Haiku 4.5Similar price
+1.0 pts vs GPT-5 Mini (predecessor) · Ahead of 3 of 4 similarly priced models
FrontierMath-Tier-4-2025-07-01-Private
Best score per model · Higher is better
#38 of 46
- 14.6%Kimi K2.6Similar price
- 6.3%GPT-5 MiniPredecessor
- 2.1%GLM 5Similar price
- 2.1%Claude Haiku 4.5Similar price
- 2.1%GPT-5 NanoNearby rank
- 2.1%o3Nearby rank
- 2.1%GPT-5.4 Mini
- 0.0%Claude Sonnet 3.5 v2Nearby rank
- 0.0%Claude Sonnet 4Nearby rank
- 0.0%Qwen3 235B A22B Thinking 2507Similar price
-4.2 pts vs GPT-5 Mini (predecessor) · Ahead of 1 of 4 similarly priced models
FrontierMath-Tier-4-v2-Private
Best score per model · Higher is better
#60 of 72
- 29.3%GLM 5.2Similar price
- 22.0%Gemini 3.8 FlashSimilar price
- 14.6%Grok 4.3Similar price
- 12.2%GPT-5 MiniPredecessor
- 12.2%GPT-5.4 NanoNearby rank
- 9.8%GPT-5.4 Mini
- 4.9%Claude Opus 4.5Nearby rank
- 4.9%InklingSimilar price
-2.4 pts vs GPT-5 Mini (predecessor) · Ahead of 1 of 4 similarly priced models
FrontierMath-Tiers-1-3-v2-Private
Best score per model · Higher is better
#53 of 87
- 68.4%Gemini 3.8 FlashSimilar price
- 59.2%GLM 5.2Similar price
- 53.0%Grok 4.7Nearby rank
- 51.2%Gemini 3 Flash PreviewNearby rank
- 51.2%GPT-5.4 Mini
- 46.7%GPT-5 MiniPredecessor
- 46.3%Inkling SmallNearby rank
- 42.8%Grok 4.3Similar price
- 33.3%InklingSimilar price
+4.6 pts vs GPT-5 Mini (predecessor) · Ahead of 2 of 4 similarly priced models
GPQA diamond
Best score per model · Higher is better
#62 of 161
- 95.4%Gemini 3.8 FlashSimilar price
- 91.9%GLM 5.2Similar price
- 88.8%Grok 4.3Similar price
- 88.3%InklingSimilar price
- 87.4%
Qwen3.6 Max PreviewNearby rank - 87.4%Claude Sonnet 4.6Nearby rank
- 86.9%GPT-5.4 Mini
- 86.4%Qwen3.5 397B A17BNearby rank
- 86.2%GPT-5Nearby rank
- 75.0%GPT-5 MiniPredecessor
+11.9 pts vs GPT-5 Mini (predecessor) · Ahead of 0 of 4 similarly priced models
OTIS Mock AIME 2024-2025
Best score per model · Higher is better
#59 of 158
- 98.9%Gemini 3.8 FlashSimilar price
- 93.3%Grok 4.3Similar price
- 91.1%GLM 5.3Nearby rank
- 90.0%Inkling SmallNearby rank
- 88.9%GPT-5.4 Mini
- 88.9%gpt-oss-120bNearby rank
- 88.9%Qwen3.5 397B A17BNearby rank
- 88.9%InklingSimilar price
- 86.7%GPT-5 MiniPredecessor
- 86.4%GLM 5.2Similar price
+2.2 pts vs GPT-5 Mini (predecessor) · Ahead of 1 of 4 similarly priced models
SimpleQA Verified
Best score per model · Higher is better
#69 of 87
- 69.7%Gemini 3.8 FlashSimilar price
- 40.3%InklingSimilar price
- 34.2%GLM 5.2Similar price
- 33.2%Grok 4.3Similar price
- 30.7%Claude Sonnet 4.5Nearby rank
- 30.2%Grok 4.20 (Reasoning)Nearby rank
- 29.4%GPT-5.4 Mini
- 26.0%GPT-4o (2024-08-06)Nearby rank
- 25.4%Qwen3.5 PlusNearby rank
- 21.6%GPT-5 MiniPredecessor
+7.8 pts vs GPT-5 Mini (predecessor) · Ahead of 0 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.
Reasoning effort
ARC-AGI-2
Score by reasoning effort · Higher is better
FrontierMath-Tiers-1-3-v2-Private
Score by reasoning effort · Higher is better
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 #871447
+58 vs GPT-5 Mini (predecessor)
- high144795% CI 1444 to 1451 · 62,795 votes
- Arena WebDevArena #1071398
- high139895% CI 1391 to 1404 · 11,817 votes
- Arena VisionArena #541250
+65 vs GPT-5 Mini (predecessor)
- high125095% CI 1244 to 1256 · 25,535 votes
- ARC-AGI-2#50 of 8518.9%
+14.5 pts vs GPT-5 Mini (predecessor)
- low1.1%
- medium4.4%
- high13.2%
- xhigh18.9%
- FrontierMath-2025-02-28-Private#16 of 5828.3%
+1.0 pts vs GPT-5 Mini (predecessor)
- high28.3%
Source:Epoch AI evaluation - FrontierMath-Tier-4-2025-07-01-Private#38 of 462.1%
-4.2 pts vs GPT-5 Mini (predecessor)
- high2.1%
Source:Epoch AI evaluation - FrontierMath-Tier-4-v2-Private#60 of 729.8%
-2.4 pts vs GPT-5 Mini (predecessor)
- xhigh9.8%
Source:Epoch AI evaluation - FrontierMath-Tiers-1-3-v2-Private#53 of 8751.2%
+4.6 pts vs GPT-5 Mini (predecessor)
- none17.2%
- low24.6%
- xhigh51.2%
Source:Epoch AI evaluation - GPQA diamond#62 of 16186.9%
+11.9 pts vs GPT-5 Mini (predecessor)
- none64.1%
- high83.6%
- xhigh86.9%
Source:Epoch AI evaluation - OTIS Mock AIME 2024-2025#59 of 15888.9%
+2.2 pts vs GPT-5 Mini (predecessor)
- none26.7%
- high87.2%
- xhigh88.9%
Source:Epoch AI evaluation - SimpleQA Verified#69 of 8729.4%
+7.8 pts vs GPT-5 Mini (predecessor)
- high29.4%
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.