Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Embodied Planning on ALFWorld Text Aligned (134 unseen tasks)
Loading...
98.5
Success Rate
EvoPlan
12.7
34.975
57.25
79.525
Jul 7, 2026
Success Rate
# Tokens
Updated 16d ago
Evaluation Results
Method
Method
Links
Success Rate
# Tokens
EvoPlan
Backbone=Qwen3-32B
2026.07
98.5
853,022
NL-PDDL
2026.07
94
443,407
Reflexion
Planner Type=Reflectiv...
2026.07
91
-
Reflexion
Planner Type=Reflectiv...
2026.07
83
-
ReAct
Planner Type=Reflectiv...
2026.07
81
5,507,266
DEPS
Planner Type=Reflectiv...
2026.07
76
-
ReAct
Planner Type=Reflectiv...
2026.07
34
4,242,560
BUTLER
Fine-tuning=100K exper...
2026.07
26
-
GPT-4o
Planner Type=Direct LL...
2026.07
21
1,358,934
LLaMA-3.1
Planner Type=Direct LL...
2026.07
19
1,193,640
Gemini-2.0 Flash
Planner Type=Direct LL...
2026.07
16
1,176,406
Feedback
Search any
task
Search any
task