Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
General Deep Research Tool Use on HLE
Loading...
42.9
Success Rate
Gemini-3-Pro
1.196
12.023
22.85
33.677
Mar 10, 2026
Success Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Success Rate
Gemini-3-Pro
Category=Frontier (≫8B...
2026.03
42.9
Gemini-2.5-Pro
Category=Frontier (≫8B...
2026.03
28.4
Kimi-K2-0905
Category=Frontier (≫8B...
2026.03
26.9
Claude-4-Sonnet
Category=Frontier (≫8B...
2026.03
20.8
GPT-OSS-120B
Category=Frontier (≫8B...
2026.03
19
DeepSeek-V3.2-Exp
Category=Frontier (≫8B...
2026.03
17.9
DIVE-8B (RL)
Category=Ours, Tempera...
2026.03
17.8
WebExplorer-8B
Category=8B Baselines,...
2026.03
17.3
DIVE-8B (SFT)
Category=Ours, Tempera...
2026.03
13.8
SWE-Dev-8B
Category=8B Baselines,...
2026.03
6.9
Qwen3-8B (base)
Category=Ours, Tempera...
2026.03
6.4
EnvScaler-8B
Category=8B Baselines,...
2026.03
2.8
Feedback
Search any
task
Search any
task