List your company

Information-seeking on GAIA

70.5Success RateOpenAI-o3
21.41234.15646.959.644Feb 11, 2026Mar 6, 2026Mar 29, 2026Apr 21, 2026May 14, 2026Jun 6, 2026Jun 30, 2026

Evaluation results

MethodLinks
OpenAI-o32026.0270.5
Claude-4-Sonnet2026.0268.3
OpenAI DeepResearch2026.0267.4
Qwen3-VL-30B-A3B-ICATools=Search & Fetch (snap.)2026.0265
MiroThinker-32B-DPO-v0.2Tools=Search & Fetch (text)2026.0264.1
Qwen3-VL-8B-ICATools=Search & Fetch (snap.)2026.0261.2
Kimi-K2-Instruct-1TTools=Search & Fetch (text)2026.0257.7
C-GRPOTools=Search & Fetch (text)2026.0256.3
WebShaper-QwQ-32BTools=Search & Fetch (text)2026.0253.3
WebSailor-32BTools=Search & Fetch (text)2026.0253.2
ASearcher-Web-32BTools=Search & Fetch (text)2026.0252.8
WebDancer-QwQ-32BTools=Search & Fetch (text)2026.0251.5
WebExplorer-8BTools=Search & Fetch (text)2026.0250
ECHO w/ Top-K retrievalBackbone=Qwen3-32B-Instruct2026.0629.1
ECHOBackbone=Qwen3-32B-Instruct2026.0629.1
GRPOBackbone=Qwen3-32B-Instruct2026.0625.2
SUPOBackbone=Qwen3-32B-Instruct2026.0625.2
ECHO w/o traceable CABackbone=Qwen3-32B-Instruct2026.0625.2
ECHOBackbone=Qwen3-30B-A3B-Instruct2026.0624.3
ECHO w/ Top-K retrieval & w/o turn summaryBackbone=Qwen3-32B-Instruct2026.0623.3
ECHO w/o Traceable CA & Turn-level ISBackbone=Qwen3-32B-Instruct2026.0623.3
GRPOBackbone=Qwen3-30B-A3B-Instruct2026.0623.3
SUPOBackbone=Qwen3-30B-A3B-Instruct2026.0623.3