WideSearch

WideSearch - agentic web search benchmark.

Category

General Agents

Max Score

100.0

Score Type

percent

Active

Yes

Model Scores

Scores for WideSearch

Model Score Date Verified
Qwen3.6-35B-A3B
48.73%
16.04.2026 Verified
Gemma4-31B-it
35.20
16.04.2026 Unverified
Qwen3.5-35B-A3B
46.77%
16.04.2026 Unverified
Gemma4-26B-A4B
6.07%
16.04.2026 Unverified
Qwen3.5-27B
61.06%
16.04.2026 Unverified
Fable-5
90.02%
18.08.2026 Unverified
Claude Opus 4.8
73.78%
18.08.2026 Unverified
Qwen3.7-Max
78.28%
18.08.2026 Unverified
Qwen3.8-2.4T-A95B
91.39%
18.08.2026 Verified
Qwen3.5-122B-A10B
49.51%
01.02.2026 Verified
GPT-5-mini 2025-08-07
23.48%
01.02.2026 Unverified
Ornith-1.5-35B-A3B
63.80%
19.08.2026 Verified
Ornith-1.0-35B-A3B
55.19%
19.08.2026 Unverified
Gemma-4-31B
37.18%
19.08.2026 Unverified
Qwen3.5-397B-A17B
75.93%
19.08.2026 Unverified
Kimi K2.6
89.24%
23.08.2026 Verified
Kimi K2.5
73.39%
23.08.2026 Verified
Hy3
91.39%
28.08.2026 Verified
Hy4 preview
95.30%
28.08.2026 Verified
DeepSeek V4 Pro
91.19%
28.08.2026 Verified
GLM-5.3
93.93%
28.08.2026 Verified
Kimi K3
89.63%
28.08.2026 Verified
GPT-5.6 Sol
100.00%
28.08.2026 Verified
Claude Opus 5
95.50%
28.08.2026 Verified