Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Web Navigation on WorkArena L1
Loading...
79.1
Success Rate
GenericAgent
-1.292
19.579
40.45
61.321
May 19, 2026
May 22, 2026
May 26, 2026
May 29, 2026
Jun 2, 2026
Jun 5, 2026
Jun 9, 2026
Success Rate
Training Speed-up
Updated 1mo ago
Evaluation Results
Method
Method
Links
Success Rate
Training Speed-up
GenericAgent
Model=GPT-5, Access=Pr...
2026.06
79.1
-
WebChallenger
Model=GLM-4-32B + Qwen...
2026.06
70.9
-
GenericAgent
Model=Claude 3.5 Sonne...
2026.06
56.4
-
Agent-as-Annotators
Model=Qwen3.5-9B, Acce...
2026.06
51.5
-
GenericAgent
Model=GPT-oss-120b, Ac...
2026.06
50.9
-
GenericAgent
Model=GPT-4o, Access=P...
2026.06
45.5
-
GenericAgent
Model=GPT-4o-mini, Acc...
2026.06
27
-
Qwen2.5-7B-Instruct + WEASEL
Training Steps=10K ste...
2026.05
7.6
9.7
Qwen2.5-7B-Instruct + Pruning + Sampling
Training Steps=10K ste...
2026.05
7
9.7
Qwen2.5-7B-Instruct + Full
Training Steps=52K ste...
2026.05
5.2
1
Qwen2.5-7B-Instruct
Zero-shot transfer=true
2026.05
4.8
-
Qwen2.5-7B-Instruct + Pruning
Training Steps=52K ste...
2026.05
1.8
1.7
Feedback
Search any
task
Search any
task