ProgramBench

Rebuilding whole programs from scratch so they pass the original behaviour tests. Run by Vals AI.

CategoryCoding
MetricAccuracy (%)
SourceVals AI (evaluator)
LicenseNone stated
Benchmark pagewww.vals.ai
StatusScored
Weight1.0 within Coding
FitDiscrimination 1.02, noise 0.63

Results

#ModelProviderResultDate
1 Claude Fable 5.2 High Anthropic 19.0% 2026-11-09
2 Claude Opus 5.5 Max Anthropic 18.5% 2026-11-09
3 Claude Fable 5.1 Max Anthropic 7.0% 2026-11-09
4 Claude Sonnet 5.5 Max Anthropic 6.5% 2026-11-09
5 GPT-6 Astra Max OpenAI 5.5% 2026-11-09
6 GPT-6.1 Sol Max OpenAI 3.0% 2026-11-09
7 Claude Opus 5 Max Anthropic 3.0% 2026-11-09
8 Gemini 4 Argon High Google 2.5% 2026-11-09
9 Claude Haiku 5.5 Max Anthropic 2.5% 2026-11-09
10 Muse Spark 1.3 Max Meta 2.5% 2026-11-09
11 Kimi K3 Moonshot AI 2.0% 2026-11-09
12 GPT-6 Sol Max OpenAI 2.0% 2026-11-09
13 Claude Fable 5 Max Anthropic 2.0% 2026-11-09
14 GLM-5.3 Max Zhipu AI 1.5% 2026-11-09
15 GPT-5.6 Sol Max OpenAI 1.5% 2026-11-09
16 Gemini 3.8 Flash High Google 1.0% 2026-11-09
17 Grok 4.6 High xAI 1.0% 2026-11-09
18 Claude Opus 4.8 Max Anthropic 1.0% 2026-11-09
19 Inkling Small Thinking Machines 0.5% 2026-11-09
20 Inkling 2 Small Thinking Machines 0.5% 2026-11-09
21 MiMo-V2.6-Flash Xiaomi 0.5% 2026-11-09
22 MiMo-V2.6-Pro Xiaomi 0.5% 2026-11-09
23 DeepSeek V4.1 Flash Max DeepSeek 0.5% 2026-11-09
24 GPT-6 Luna Max OpenAI 0.5% 2026-11-09
25 Grok 4.7 Extra High xAI 0.5% 2026-11-09
26 Claude Sonnet 4.6 Max Anthropic 0.5% 2026-11-09
27 GPT-5.4 High OpenAI 0.5% 2026-11-09
28 GPT-5.6 Terra Max OpenAI 0.5% 2026-11-09
29 GLM-5.2 Max Zhipu AI 0.5% 2026-11-09
30 GPT-5.5 Extra High OpenAI 0.5% 2026-11-09
31 Muse Spark 1.3 Extra High Meta 0.5% 2026-11-09
32 Muse Spark 1.2 Extra High Meta 0.5% 2026-11-09
33 Muse Wonder Full Max Meta 0.5% 2026-11-09
34 Nemotron 3 Ultra 550B A55B NVIDIA 0.0% 2026-11-09
35 Kimi K2.7 Code Moonshot AI 0.0% 2026-11-09
36 Kimi K2.6 Moonshot AI 0.0% 2026-11-09
37 Hy4 preview Tencent 0.0% 2026-11-09
38 MiniMax-M2.7 MiniMax 0.0% 2026-11-09
39 GLM-5.1 Zhipu AI 0.0% 2026-11-09
40 Laguna XS.2 Poolside 0.0% 2026-11-09
41 Laguna M.1 Poolside 0.0% 2026-11-09
42 Inkling Thinking Machines 0.0% 2026-11-09
43 Qwen3.8 Max Alibaba 0.0% 2026-11-09
44 Qwen3.6 Plus Alibaba 0.0% 2026-11-09
45 GLM-5.3-Flash Max Zhipu AI 0.0% 2026-11-09
46 DeepSeek V4 Pro 0813 Max DeepSeek 0.0% 2026-11-09
47 Gemini 3.7 Flash High Google 0.0% 2026-11-09
48 GPT-5.4 mini Extra High OpenAI 0.0% 2026-11-09
49 Gemini 3.5 Flash Lite High Google 0.0% 2026-11-09
50 Gemini 3.1 Pro Preview High Google 0.0% 2026-11-09
51 Gemini 3 Flash Preview High Google 0.0% 2026-11-09
52 Claude Sonnet 5 Max Anthropic 0.0% 2026-11-09
53 Claude Opus 4.7 Max Anthropic 0.0% 2026-11-09
54 Gemini 3.6 Flash High Google 0.0% 2026-11-09
55 GPT-5.6 Luna Max OpenAI 0.0% 2026-11-09
56 Grok 4.5 High xAI 0.0% 2026-11-09
57 Grok 4.3 High xAI 0.0% 2026-11-09
58 Gemini 3.5 Flash High Google 0.0% 2026-11-09
59 GPT-5.4 Extra High OpenAI 0.0% 2026-11-09
60 Claude Haiku 4.5 Thinking Anthropic 0.0% 2026-11-09
61 Muse Spark 1.1 Extra High Meta 0.0% 2026-11-09
62 DeepSeek V4 Flash 0731 High DeepSeek 0.0% 2026-11-09
63 Qwen3.8 27B Extra High Alibaba 0.0% 2026-11-09
64 DeepSeek V4 Pro Max DeepSeek 0.0% 2026-11-09
65 Gemini 3.1 Flash Lite Preview High Google 0.0% 2026-11-09