Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Agentic Coding on TerminalBench 2
Loading...
81.8
Pass Rate
ForgeCode
11.184
29.517
47.85
66.183
Mar 30, 2026
Pass Rate
Updated 18d ago
Evaluation Results
Method
Method
Links
Pass Rate
ForgeCode
Base Model=Claude Opus...
2026.03
81.8
Meta-Harness
Base Model=Claude Opus...
2026.03
76.4
Capy
Base Model=Claude Opus...
2026.03
75.3
Terminus-KIRA
Base Model=Claude Opus...
2026.03
74.7
MAYA-V2
Base Model=Claude Opus...
2026.03
72.1
TongAgents
Base Model=Claude Opus...
2026.03
71.9
Droid
Base Model=Claude Opus...
2026.03
69.9
Mux
Base Model=Claude Opus...
2026.03
66.5
Terminus 2
Base Model=Claude Opus...
2026.03
62.9
Claude Code
Base Model=Claude Opus...
2026.03
58
Meta-Harness
Base Model=Claude Haik...
2026.03
37.6
Goose
Base Model=Claude Haik...
2026.03
35.5
Terminus-KIRA
Base Model=Claude Haik...
2026.03
33.7
Mini-SWE-Agent
Base Model=Claude Haik...
2026.03
29.8
Terminus 2
Base Model=Claude Haik...
2026.03
28.3
Claude Code
Base Model=Claude Haik...
2026.03
27.5
OpenHands
Base Model=Claude Haik...
2026.03
13.9
Feedback
Search any
task
Search any
task