// BENCHMARK

GAIA benchmark

AI model leaderboard for the GAIA benchmark. Compare how large language models score on GAIA, see the full ranking, and understand what this AI benchmark measures. Tongyi DeepResearch currently leads with 70.9. General AI Assistants benchmark — multi-step real-world tasks requiring web search, code, and reasoning.

Leaderboard

# Model Organization Score Variant Source
#1 Tongyi DeepResearch Alibaba-NLP 70.9 react-avg@3 official ↗
#2 Claude Sonnet 4 Anthropic 68.3 react-cited-tongyi official ↗
#3 GLM 4.5 Zhipu AI 66 react-cited-tongyi official ↗
#4 DeepSeek-V3.1 DeepSeek 63.1 react-cited-tongyi official ↗
#5 o4-mini OpenAI 60 react-cited-tongyi official ↗
#6 Kimi K2 Moonshot 57.7 react-cited-tongyi official ↗

Frequently asked questions about GAIA

What is the GAIA benchmark?

General AI Assistants benchmark — multi-step real-world tasks requiring web search, code, and reasoning.

How is the GAIA benchmark scored?

GAIA is scored using the pass@1 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 GAIA?

As of the latest reported scores on GenAIList, Tongyi DeepResearch achieves the highest result on GAIA with a score of 70.9.

Is a higher GAIA score better?

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