List your company

General AI Assistant on GAIA text

70.5GAIA Average ScoreOpenAI-o3
15.3829.694458.31Jan 29, 2026

Evaluation results

MethodLinks
OpenAI-o32026.0170.5———
Claude-4-Sonnet2026.0168.3———
OpenAI DeepResearch2026.0167.4———
ARPOBackbone=Qwen3-14B2026.0143.756.440.416.7
Reagent-UBackbone=Qwen3-8B2026.0143.75938.516.7
WebDancerBackbone=Qwen2.5-32B2026.0140.746.144.28.3
Search-o1Backbone=QwQ-32B-Preview2026.0139.853.834.616.7
ARPOBackbone=Qwen3-8B2026.0138.853.932.716.7
Reagent-RBackbone=Qwen3-8B2026.0136.951.330.816.7
VerlToolBackbone=Qwen3-8B2026.0134———
Reagent w/o Agent-RRMBackbone=Qwen3-8B2026.01344136.50
WebDancerBackbone=Qwen2.5-7B2026.01314130.70
DeepSeek-R1-671BBackbone=DeepSeek-R1-671B2026.0125.240.521.25.2
Reagent-CBackbone=Qwen3-8B, Inference Type=Direct Inference2026.0125.230.823.116.7
WebThinkerBackbone=Qwen3-8B2026.0122.343.611.50
Qwen3-8BBackbone=Qwen3-8B2026.0121.428.221.20
Tree-GRPOBackbone=Qwen2.5-14B2026.012120.824.37.3
QwQ-32BBackbone=QwQ-32B2026.0118.930.96.55.2
Atom-SearcherBackbone=Qwen2.5-7B2026.0117.51821.20