Inference
Vals AI

SkillsBench

How important are skills for agents?

ModelsModels with a result.34
Top resultBest result on this test.69.8%DeepSeek V4.1 Flash
Top-3 spreadPoints from first to third.3.9 pts
UpdatedDate the source changed the results.11 Sept 2026

Result and cost

020406080100$0.03$0.1$0.3$1$3$10RESULTUS DOLLARS PER TASK · LOG SCALE

Results

34 results
#ModelResultThe score from the source.CostUS dollars to run one task.TimeTime to run one task.
01DeepSeek V4.1 FlashOpenHands69.8%$0.0854 min
02Grok 4.5OpenHands66.0%$0.5782 min
03Gemini 3.7 FlashOpenHands65.9%$1.83 min
04GPT-5.5 CodexCodex62.6%$1.575 min
05GPT-5.5OpenHands62.2%$2.547 min
06Claude Fable 5.1OpenHands61.6%$4.5211 min
07GPT-5.6 LunaOpenHands60.4%$0.165 min
08Claude Opus 5OpenHands60.4%$2.518 min
09Claude Opus 4.8OpenHands59.2%$4.7716 min
10Muse Spark 1.1OpenHands59.2%$0.4496 min
11GPT-5.6 TerraOpenHands58.9%$1.47 min
12Gemini 3.8 FlashOpenHands58.0%$2.344 min
13Grok 4.6OpenHands55.8%$0.92113 min
14Qwen3.7 PlusOpenHands54.3%$0.13712 min
15GPT-5.6 SolOpenHands54.1%$4.1413 min
16Muse Spark 1.2OpenHands53.0%$0.64913 min
17Gemini 3.5 FlashOpenHands52.7%$1.257 min
18DeepSeek V4 Pro 0423OpenHands52.6%$0.37511 min
19GPT-5.4OpenHands51.7%$2.049 min
20MiniMax M3OpenHands51.5%$0.50515 min
21DeepSeek V4 Flash 0423OpenHands50.7%$0.1267 min
22Kimi K2.7 CodeOpenHands50.0%$0.44313 min
23Claude Sonnet 4.6OpenHands49.1%$1.3814 min
24GLM-5.3OpenHands47.5%$0.71413 min
25Claude Sonnet 5OpenHands46.5%$4.3515 min
26GLM-5.2OpenHands45.1%$0.81120 min
27Qwen3.8 MaxOpenHands42.0%$0.52516 min
28Grok 4.3OpenHands40.6%$0.425 min
29GLM-5.3-FlashOpenHands40.2%$0.03912 min
30Qwen3.8-27BOpenHands38.1%$0.8187 min
31Inkling-SmallOpenHands33.6%$0.14812 min
32Ling 3.0 FlashOpenHands27.9%$0.0615 min
33InklingOpenHands26.1%$0.82716 min
34Mercury 2.5OpenHands18.1%$0.10149 s