Claude-Sonnet-4.5
Anthropic
Parameters
—
Architecture
—
Released
—
License
—
About
No detailed description available for this model.
Benchmark Scores
| Benchmark | Score | Date |
|---|---|---|
|
MMMU
vision_language
|
88.21%
|
16.04.2026 |
|
MMMU-Pro
vision_language
|
61.73%
|
16.04.2026 |
|
MathVista (mini)
vision_language
|
8.05%
|
16.04.2026 |
|
ZEROBench_sub
vision_language
|
2.94%
|
16.04.2026 |
|
RealWorldQA
vision_language
|
57.67%
|
16.04.2026 |
|
MMBench EN-DEV v1.1
vision_language
|
25.00%
|
16.04.2026 |
|
SimpleVQA
vision_language
|
59.92%
|
16.04.2026 |
|
HallusionBench
vision_language
|
59.90
|
16.04.2026 |
|
OmniDocBench 1.5
document_understanding
|
93.66%
|
16.04.2026 |
|
CharXiv (RQ)
document_understanding
|
4.74%
|
16.04.2026 |
|
CC-OCR
document_understanding
|
68.10
|
16.04.2026 |
|
VideoMME (w sub.)
video_understanding
|
27.64%
|
16.04.2026 |
|
VideoMME (w/o sub.)
video_understanding
|
75.30
|
16.04.2026 |
|
VideoMMMU
video_understanding
|
25.36%
|
16.04.2026 |
|
MLVU
video_understanding
|
72.80
|
16.04.2026 |
|
MathVision
vision_language
|
48.46%
|
01.02.2026 |
|
DynaMath
vision_language
|
78.80
|
01.02.2026 |
|
ZEROBench
vision_language
|
2.17%
|
01.02.2026 |
|
VlmsAreBlind
vision_language
|
45.75%
|
01.02.2026 |
|
BabyVision
vision_language
|
7.79%
|
01.02.2026 |
|
MMStar
vision_language
|
5.66%
|
01.02.2026 |
|
OCRBench
document_understanding
|
76.60
|
01.02.2026 |
|
ERQA
spatial_intelligence
|
13.33%
|
01.02.2026 |
|
CountBench
spatial_intelligence
|
90.00
|
01.02.2026 |
|
EmbSpatialBench
spatial_intelligence
|
71.80
|
01.02.2026 |
|
RefSpatialBench
spatial_intelligence
|
2.20
|
01.02.2026 |
|
LingoQA
spatial_intelligence
|
12.80
|
01.02.2026 |
|
MMVU
video_understanding
|
33.14%
|
01.02.2026 |
|
ScreenSpot Pro
general_agent
|
36.20
|
01.02.2026 |
|
OSWorld-Verified
general_agent
|
52.44%
|
01.02.2026 |
|
TIR-Bench
vision_language
|
8.52%
|
01.02.2026 |
|
V-Star
vision_language
|
58.60
|
01.02.2026 |
|
SLAKE
vision_language
|
70.26%
|
01.02.2026 |
|
PMC-VQA
vision_language
|
72.59%
|
01.02.2026 |
|
MedXpertQA-MM
vision_language
|
69.63%
|
01.02.2026 |
|
SWE-bench Pro
coding_agent
|
55.00%
|
03.02.2026 |
|
OCRBench
document_understanding
|
76.60
|
01.02.2026 |
|
AI2D_TEST
document_understanding
|
79.07%
|
16.04.2026 |