Claude Opus 4.6
Anthropic
Parameters
—
Architecture
—
Released
—
License
—
About
No detailed description available for this model.
Benchmark Scores
| Benchmark | Score | Date | Standing |
|---|---|---|---|
|
AgentWorldBench MCP
general_agent
|
98.68%
|
24.06.2026 |
#3/14
better than 85%
|
|
AgentWorldBench Search
general_agent
|
46.35%
|
24.06.2026 |
#7/14
better than 54%
|
|
AgentWorldBench Terminal
general_agent
|
90.49%
|
24.06.2026 |
#3/14
better than 85%
|
|
AgentWorldBench SWE
general_agent
|
87.31%
|
24.06.2026 |
#4/14
better than 77%
|
|
AgentWorldBench Android
general_agent
|
100.00%
|
24.06.2026 |
#1/14
better than 100%
|
|
AgentWorldBench Web
general_agent
|
57.14%
|
24.06.2026 |
#5/14
better than 69%
|
|
AgentWorldBench OS
general_agent
|
100.00%
|
24.06.2026 |
#1/14
better than 100%
|
|
AgentWorldBench Overall
general_agent
|
92.77%
|
24.06.2026 |
#3/14
better than 85%
|
|
MMMU-Pro
vision_language
|
82.32%
|
23.08.2026 |
#21/35
better than 41%
|
|
HLE (with tools)
stem_reasoning
|
69.70%
|
23.08.2026 |
#22/49
better than 56%
|
|
DeepSearch QA
general_agent
|
88.89%
|
23.08.2026 |
#3/9
better than 75%
|
|
Toolathlon Verified
general_agent
|
38.72%
|
23.08.2026 |
#28/35
better than 18%
|
|
MCPMark
general_agent
|
59.00%
|
23.08.2026 |
#3/15
better than 86%
|
|
Claw-Eval Avg
coding_agent
|
100.00%
|
23.08.2026 |
#1/40
better than 100%
|
|
Apex-Agents
general_agent
|
42.68%
|
23.08.2026 |
#15/24
better than 35%
|
|
OSWorld-Verified
general_agent
|
75.41%
|
23.08.2026 |
#11/25
better than 54%
|
|
Terminal-Bench 2.0
coding_agent
|
91.53%
|
23.08.2026 |
#4/23
better than 77%
|
|
SWE-bench Multilingual
coding_agent
|
86.15%
|
23.08.2026 |
#7/48
better than 87%
|
|
SWE-bench Verified
coding_agent
|
92.20%
|
23.08.2026 |
#5/60
better than 92%
|
|
SciCode (subtask)
stem_reasoning
|
66.02%
|
23.08.2026 |
#5/15
better than 71%
|
|
OJBench
stem_reasoning
|
84.73%
|
23.08.2026 |
#3/22
better than 90%
|
|
LiveCodeBench v6
stem_reasoning
|
93.59%
|
23.08.2026 |
#12/49
better than 71%
|
|
Humanity's Last Exam
stem_reasoning
|
71.78%
|
23.08.2026 |
#19/71
better than 71%
|
|
AIME 26
stem_reasoning
|
96.81%
|
23.08.2026 |
#10/45
better than 80%
|
|
HMMT Feb 26
stem_reasoning
|
98.06%
|
23.08.2026 |
#5/44
better than 88%
|
|
IMOAnswerBench
stem_reasoning
|
74.86%
|
23.08.2026 |
#19/30
better than 34%
|
|
GPQA Diamond
stem_reasoning
|
94.33%
|
23.08.2026 |
#17/80
better than 76%
|
|
CharXiv (RQ)
document_understanding
|
12.25%
|
23.08.2026 |
#24/34
better than 30%
|
|
MathVision
vision_language
|
48.65%
|
23.08.2026 |
#20/27
better than 27%
|
|
V-Star
vision_language
|
69.85%
|
23.08.2026 |
#11/20
better than 42%
|
|
BrowseComp
general_agent
|
90.56%
|
23.08.2026 |
#6/45
better than 86%
|
|
SWE-bench Pro
coding_agent
|
59.53%
|
23.08.2026 |
#40/77
better than 46%
|
|
Claw-Eval Pass^3
coding_agent
|
85.18%
|
23.08.2026 |
#2/21
better than 95%
|
|
BabyVision
vision_language
|
2.70%
|
23.08.2026 |
#23/24
better than 4%
|
What is a good AgentWorldBench MCP score? What is a good AgentWorldBench Search score? What is a good AgentWorldBench Terminal score?
No comments yet
Be the first to share your thoughts!