// BENCHMARK

BBEH benchmark

AI model leaderboard for the BBEH benchmark. Compare how large language models score on BBEH, see the full ranking, and understand what this AI benchmark measures. GPT-5 currently leads with 72.78. BIG-Bench Extra Hard — successor to BBH that probes deeper reasoning on harder, more diverse tasks.

Leaderboard

# Model Organization Score Variant Source
#1 GPT-5 OpenAI 72.78 thinking-high-cited-ring-1t official ↗
#2 DeepSeek-V3.1-Terminus DeepSeek 61.04 thinking-cited-ring-1t official ↗
#3 Qwen3-235B-A22B-Thinking (Jul 2025) Qwen 60 thinking-cited-ring-1t official ↗
#4 Ring-1T Ant Group 59.63 thinking official ↗
#5 gemini-2.5-pro Google DeepMind 51.51 cited-ring-1t official ↗
#6 Ling-1T Ant Group 47.34 official ↗
#7 DeepSeek-V3.1-Terminus DeepSeek 42.86 non-thinking-cited-ling-1t official ↗
#8 GPT-5 OpenAI 39.75 main-cited-ling-1t official ↗
#9 Kimi K2 Moonshot 34.83 instruct-0905-cited-ling-1t official ↗
#10 gemini-2.5-pro Google DeepMind 29.08 lowthink-cited-ling-1t official ↗

Frequently asked questions about BBEH

What is the BBEH benchmark?

BIG-Bench Extra Hard — successor to BBH that probes deeper reasoning on harder, more diverse tasks.

How is the BBEH benchmark scored?

BBEH is scored using the accuracy metric, where a higher score is better. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.

Which AI model scores highest on BBEH?

As of the latest reported scores on GenAIList, GPT-5 achieves the highest result on BBEH with a score of 72.78.

Is a higher BBEH score better?

Yes. On BBEH a higher score indicates better performance, so models near the top of the leaderboard are the strongest.