Claude-Opus-4.6 (Max)
Anthropic
Parameters
—
Architecture
Proprietary (unknown)
Released
26.08.2026
License
—
Input Modalities
Output Modalities
About
No detailed description available for this model.
Benchmark Scores
| Benchmark | Score | Date | Standing |
|---|---|---|---|
|
HLE (with tools)
stem_reasoning
|
46.52%
|
26.08.2026 |
#37/49
better than 25%
|
|
SWE-bench Multilingual
coding_agent
|
85.80%
|
26.08.2026 |
#8/48
better than 81%
|
|
CoWorkBench
general_agent
|
75.00%
|
26.08.2026 |
#7/12
better than 36%
|
|
GPQA Diamond
stem_reasoning
|
94.33%
|
26.08.2026 |
#17/80
better than 76%
|
|
LiveCodeBench v6
stem_reasoning
|
93.59%
|
26.08.2026 |
#12/49
better than 71%
|
|
AndroidWorld
general_agent
|
13.13%
|
26.08.2026 |
#11/13
better than 8%
|
|
RealWorldQA
vision_language
|
66.05%
|
26.08.2026 |
#15/23
better than 32%
|
|
CharXiv (RQ)
document_understanding
|
66.00
|
26.08.2026 |
#33/34
better than 0%
|
|
Claw-Eval Avg
coding_agent
|
65.24%
|
26.08.2026 |
#27/40
better than 28%
|
|
JobBench
general_agent
|
32.03%
|
26.08.2026 |
#14/19
better than 28%
|
|
MathVision
vision_language
|
37.64%
|
26.08.2026 |
#22/27
better than 15%
|
|
LVBench
video_understanding
|
61.58%
|
26.08.2026 |
#11/12
better than 9%
|
|
NL2Repo-Bench
coding_agent
|
48.42%
|
26.08.2026 |
#14/23
better than 41%
|
|
ERQA
spatial_intelligence
|
6.80%
|
26.08.2026 |
#19/21
better than 5%
|
|
SWE-bench Pro
coding_agent
|
59.53%
|
26.08.2026 |
#40/77
better than 46%
|
|
IFBench
instruction_following
|
64.72%
|
26.08.2026 |
#33/42
better than 22%
|
|
Claw-Eval Pass^3
coding_agent
|
51.59%
|
26.08.2026 |
#11/21
better than 40%
|
What is a good HLE (with tools) score? What is a good SWE-bench Multilingual score? What is a good CoWorkBench score?
No comments yet
Be the first to share your thoughts!