Support

Settings

Theme

Explore themes

Mode

Library

Looks borrowed from benchmark sites and products. Themes with a light and a dark version follow your mode.

Bring your own theme

Upload a Priors theme file, or have your agent make one: copy the prompt, tell it the look you want, and upload what it returns. With the Priors skill installed it already knows the format.

Custom themes stay in this browser. They travel inside shared chart links and a page's theme, so others see them as you do.

For the builder, Explore and model pages on this device. Published pages keep the theme their owner picked.

Mode

Benchmarks · coding

ProgramBench v1 Fully Resolved

Share of public ProgramBench v1 tasks where the implementation passes every hidden behavioral test.

53 models · 54 results · updated 2026-11-05 · vals.ai ↗

Leaderboard

Chart settings
53 of 53 selected
Opus 5.5 max

Score vs subscription cost per task

Highlight models

Click a point on the chart to highlight that exact effort; click it again to remove it.

Chart settings
53 of 53 selected

Heavy line: best score at each price. Thin lines join one model's effort levels. Cost uses each model's median subscription plan, or API price with no plan data. 33 more models under All models in settings.

More from this benchmark

46 published measures

Breakdown

darker is better within each column · click a column to sort
Model
Opus 5.5 max25655.3M124.2M25469.8M185.4M
Sonnet 5.5 max26746.1M138.9M26556.7M189.3M
GPT-6 Astra max751.2M19.6M9.2M693.9M
GPT-6.1 Sol max1004.1M30.4M18.1M950.1M54.0M
Gemini 4 Argon high5363.0M40.8M18.7M4930.4M0
Muse Spark 1.3 max10898.5M59.1M24.7M2731.6M0
GPT-6 Sol max6137.6M31.9M13.9M6073.5M63.2M
Kimi K3 max22310.4M87.2M54.3M
GLM 5.3 max15021.5M68.4M45.5M14859.3M0
Gemini 3.8 Flash high8981.1M33.8M15.1M8555.7M0
Grok 4.6 high2164.9M62.7M44.3M902.6M0
DeepSeek V4.1 Flash max14371.7M70.1M46.7M14340.4M0
GPT-5.6 Terra max3001.7M22.2M9.5M2888.8M
GPT-6 Luna max1373.3M27.1M16.7M1324.5M47.9M
Grok 4.7 xhigh4016.8M40.4M21.7M1901.6M0
Inkling 2 Small max50784.2M120.5M50607.8M0
MiMo V2.6 Flash high15388.9M107.7M80.1M15236.8M0
MiMo V2.6 Pro max8121.2M81.7M62.8M8016.9M0
Muse Spark 1.2 xhigh2012.7M28.0M14.1M1931.2M0
DeepSeek V4 Flash 0731 high11633.5M60.4M37.8M11608.8M0
DeepSeek V4 Pro 0813 max11180.5M36.7M17.1M11150.3M0
Gemini 3.7 Flash high5272.5M33.0M9.6M5006.3M0
GLM 5.3 Flash max8901.5M41.7M22.3M8651.9M0
Hy4 preview high8424.2M55.4M38.2M5827.5M0
Inkling max4102.8M18.6M4065.5M0

All results

54 rows
#ModelEffortScoreSubscription $ / taskAPI $ / taskHarness
1Opus 5.5max18.5%$1.03$68mini-SWE-agent
2Fable 5.1max7.0%$9.12$59mini-SWE-agent
3Sonnet 5.5max6.5%$0.54$36mini-SWE-agent
4GPT-6 Astramax5.5%$0.50$12mini-SWE-agent
5Opus 5max3.0%$1.04$60mini-SWE-agent
6GPT-6.1 Solmax3.0%$0.17$2.87mini-SWE-agent
7Haiku 5.5max2.5%$0.14$6.83mini-SWE-agent
8Gemini 4 Argonhigh2.5%$13$18mini-SWE-agent
9Muse Spark 1.3max2.5%$27$54mini-SWE-agent
10Fable 5max2.0%$8.53$76mini-SWE-agent
11GPT-6 Solmax2.0%$0.36$12mini-SWE-agent
12Kimi K3max2.0%$6.41$70mini-SWE-agent
13GLM 5.3max1.5%$2.28$22mini-SWE-agent
14GPT-5.6 Solmax1.5%$0.42$15mini-SWE-agent
15Opus 4.8max1.0%$0.54$31mini-SWE-agent
16Gemini 3.8 Flashhigh1.0%$0.44$11mini-SWE-agent
17Grok 4.6high1.0%$1.72$17mini-SWE-agent
18Sonnet 4.6max0.5%mini-SWE-agent
19DeepSeek V4.1 Flashmax0.5%$0.064$0.90mini-SWE-agent
20GLM 5.2max0.5%$0.37$13mini-SWE-agent
21GPT-5.5xhigh0.5%$0.19$6.95mini-SWE-agent
22GPT-5.6 Terramax0.5%$0.15$5.38mini-SWE-agent
23GPT-6 Lunamax0.5%$0.013$0.18mini-SWE-agent
24Grok 4.7xhigh0.5%$4.32$46mini-SWE-agent
25Inkling 2 Smallmax0.5%$8.46$8.46mini-SWE-agent
26Inkling Smallmax0.5%$0.39$0.48mini-SWE-agent
27MiMo V2.6 Flashhigh0.5%free$0.47mini-SWE-agent
28MiMo V2.6 Promax0.5%$0.61$0.73mini-SWE-agent
29Muse Spark 1.2xhigh0.5%$1.23$2.47mini-SWE-agent
30Muse Spark 1.3xhigh0.5%$5.78$12mini-SWE-agent