Skip to content
Weights Not Walls
Menu

CL-bench Life

Reported

No description yet: we only describe an eval set when we can do it correctly.

Kind
Published by a third party's leaderboard, collected by Epoch AI. The leaderboard is named and linked.
Score shown
Epoch's column "Overall", as a percentage, higher is better; random guessing scores about 0.0%. The source gives no error bars.
Results
17 in all: 6 for open-weight models, 10 for closed ones, 1 where Epoch does not say. 6 linked to a model page here.

Best open-weight result: deepseek-ai/DeepSeek-V4-Pro, 13.5%, number 9 of 17 results.

CL-bench Life results
#ModelWeightsSettingScoreDateSource
1gpt-5.5_highOpenAIclosedeffort high22.2%not datedvia Epoch AI
2gpt-5.4-2026-03-05_xhighOpenAIclosedeffort xhigh21.7%not datedvia Epoch AI
3gpt-5.4-2026-03-05_highOpenAIclosedeffort high19.3%not datedvia Epoch AI
4gpt-5.1-2025-11-13_highOpenAIclosedeffort high17.3%not datedvia Epoch AI
5claude-opus-4-6_highAnthropicclosedeffort high17.0%not datedvia Epoch AI
6gemini-3.1-pro-previewGoogle DeepMindcloseddefault16.9%not datedvia Epoch AI
7gpt-5.4-2026-03-05_unknownOpenAIclosedeffort unknown13.8%not datedvia Epoch AI
8claude-opus-4-6_unknownAnthropicclosedeffort unknown13.6%not datedvia Epoch AI
9deepseek-ai/DeepSeek-V4-ProEpoch: deepseek-v4-pro_high, DeepSeekopeneffort high13.5%not datedvia Epoch AI
10moonshotai/Kimi-K2.5Epoch: kimi-k2.5, Moonshotopendefault13.2%not datedvia Epoch AI
11qwen3.5-plusAlibabanot stateddefault12.4%not datedvia Epoch AI
12grok-4-20xAIcloseddefault11.9%not datedvia Epoch AI
13zai-org/GLM-4.7Epoch: zai-org/glm-4.7, Z.ai (Zhipu AI)opendefault10.9%not datedvia Epoch AI
14deepseek-ai/DeepSeek-V3.2-ExpEpoch: DeepSeek-V3.2-Exp_thinking, DeepSeekopeneffort thinking9.5%not datedvia Epoch AI
15deepseek-ai/DeepSeek-V3.2Epoch: deepseek/deepseek-v3.2, DeepSeekopendefault7.4%not datedvia Epoch AI
16mimo-v2-proXiaomi Corpcloseddefault6.9%not datedvia Epoch AI
17MiniMaxAI/MiniMax-M2.5Epoch: MiniMax-M2.5, MiniMaxopendefault6.3%not datedvia Epoch AI

Open-weight rows are shaded. Dates are when the result was run or evaluated, where the data says. Data from Epoch AI (CC-BY 4.0), fetched 19 Nov 2026: Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Epoch AI. Method