Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Deep Search on xBench DeepSearch DS-2505
Loading...
82
Score
SMTL-30B-300
52.568
60.209
67.85
75.491
Mar 30, 2026
Score
Updated 18d ago
Evaluation Results
Method
Method
Links
Score
SMTL-30B-300
Model Category=Trained...
2026.03
82
Marco-DR-8B
Model Category=Trained...
2026.03
82
DeepSeek-V3.2
Model Category=Foundat...
2026.03
78
OpenAI GPT-5 High
Model Category=Foundat...
2026.03
77.8
MiroThinker-v1.0-72B
Model Category=Trained...
2026.03
77.8
MiroThinker-v1.5-235B
Model Category=Trained...
2026.03
77.1
Tongyi-DR-30B
Model Category=Trained...
2026.03
75
OpenSeeker-30B-SFT
Model Category=Trained...
2026.03
74
RE-TRAC-4B
Model Category=Trained...
2026.03
74
WebSailor-V2-30B
Model Category=Trained...
2026.03
73.7
MiroThinker-v1.5-30B
Model Category=Trained...
2026.03
73.1
GLM-4.7
Model Category=Foundat...
2026.03
72
MiroThinker-v1.0-30B
Model Category=Trained...
2026.03
70.6
AgentCPM-Explore-4B
Model Category=Trained...
2026.03
70
Minimax-M2.1
Model Category=Foundat...
2026.03
68.7
OpenAI-o3
Model Category=Foundat...
2026.03
67
Claude-4-Sonnet
Model Category=Foundat...
2026.03
64.6
DeepMiner-32B-RL
Model Category=Trained...
2026.03
62
MiroThinker-v1.0-8B
Model Category=Trained...
2026.03
60.6
WebExplorer-8B-RL
Model Category=Trained...
2026.03
53.7
Feedback
Search any
task
Search any
task