LiveBench 2.0 Coding

Real merged-PR bug fixes on a frozen monorepo; deterministic grading, a regression scores 0. LiveBench 2.0, refreshed with new questions each release to limit contamination.

CategoryCoding
MetricScore (%)
SourceLiveBench (original)
LicenseNone stated
Benchmark pagelivebench.ai
StatusScored
Weight1.0 within Coding
FitDiscrimination 0.77, noise 0.18

Results

#ModelProviderResultDate
1 Gemini 4 Carbon Google 91.0% 2026-11-06
2 Claude Fable 5.1 Anthropic 89.7% 2026-11-06
3 GPT-6.1 Terra OpenAI 89.2% 2026-11-06
4 GPT-6 Astra OpenAI 87.5% 2026-11-06
5 Claude Opus 5.5 Anthropic 87.1% 2026-11-06
6 Gemini 4 Argon Google 85.7% 2026-11-06
7 GPT-5.6 Terra OpenAI 84.2% 2026-11-06
8 Muse Spark 1.3 Meta 83.1% 2026-11-06
9 GPT-6.1 Sol OpenAI 82.1% 2026-11-06
10 DeepSeek V4.1 Flash 1017 DeepSeek 80.4% 2026-11-06
11 Mistral Large 4 Preview Mistral 79.4% 2026-11-06
12 Claude Fable 5.2 Anthropic 79.2% 2026-11-06
13 Claude Haiku 5.5 Anthropic 77.6% 2026-11-06
14 Claude Sonnet 5.5 Anthropic 77.2% 2026-11-06
15 Grok 4.6 xAI 76.7% 2026-11-06
16 Muse Wonder Meta 74.9% 2026-11-06
17 Grok 4.7 xAI 73.1% 2026-11-06
18 GLM-5.3-Flash Zhipu AI 70.9% 2026-11-06
19 GPT-6 Luna OpenAI 70.8% 2026-11-06
20 Step 5 Preview StepFun 69.4% 2026-11-06
21 Qwen3.8 27B Alibaba 69.3% 2026-11-06
22 MiniMax-M3.1-Flash MiniMax 68.8% 2026-11-06
23 Kimi K3.1 Pro Moonshot AI 67.8% 2026-11-06
24 Gemini 3.8 Flash Google 65.7% 2026-11-06
25 MiMo-V2.6-Pro Xiaomi 64.5% 2026-11-06
26 Qwen3.8 Max Alibaba 60.9% 2026-11-06
27 Muse Spark Open Meta 55.1% 2026-11-06
28 Nemotron 3.5 Super VL 120B A12B NVIDIA 54.1% 2026-11-06
29 Kimi K3 Moonshot AI 53.1% 2026-11-06
30 Inkling Thinking Machines 48.7% 2026-11-06
31 Mistral Large 4 Mistral 44.4% 2026-11-06
32 Gemma 4 31B IT Google 20.3% 2026-11-06