Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Deep search on HLE
Loading...
11.4
Accuracy
ECHO
3.704
5.702
7.7
9.698
Jan 26, 2026
Feb 20, 2026
Mar 18, 2026
Apr 13, 2026
May 9, 2026
Jun 4, 2026
Jun 30, 2026
Accuracy
Tools Usage
Updated 24d ago
Evaluation Results
Method
Method
Links
Accuracy
Tools Usage
ECHO
Backbone=Qwen3-32B-Ins...
2026.06
11.4
-
ECHO w/ Top-K retrieval
Backbone=Qwen3-32B-Ins...
2026.06
10.6
-
ECHO w/ Top-K retrieval & w/o turn summary
Backbone=Qwen3-32B-Ins...
2026.06
10
-
ECHO w/o Traceable CA & Turn-level IS
Backbone=Qwen3-32B-Ins...
2026.06
9.4
-
SUPO
Backbone=Qwen3-32B-Ins...
2026.06
9.2
-
ECHO
Backbone=Qwen3-30B-A3B...
2026.06
9.2
-
GRPO
Backbone=Qwen3-32B-Ins...
2026.06
8.8
-
ECHO w/o traceable CA
Backbone=Qwen3-32B-Ins...
2026.06
8.8
-
Musique
Backbone=QWEN-7B, Trai...
2026.01
8
-
SUPO
Backbone=Qwen3-30B-A3B...
2026.06
8
-
GRPO
Backbone=Qwen3-30B-A3B...
2026.06
7.8
-
SAGE
Backbone=QWEN-7B, Trai...
2026.01
7
-
SAGE
Backbone=QWEN-3B, Trai...
2026.01
5.5
-
NQ + HotpotQA
Backbone=QWEN-3B, Trai...
2026.01
5
-
NQ + HotpotQA
Backbone=QWEN-7B, Trai...
2026.01
4.5
-
Musique
Backbone=QWEN-3B, Trai...
2026.01
4
-
Feedback
Search any
task
Search any
task