DeepSeek-V4-Vision-Exp
DeepSeek
Parameters
—
Architecture
MoE
Released
—
License
—
Input Modalities
text
image
Output Modalities
text
About
No detailed description available for this model.
Benchmark Scores
| Benchmark | Score | Date | Standing |
|---|---|---|---|
|
Terminal Bench 2.1
coding_agent
|
92.57%
|
27.08.2026 |
#13/44
better than 70%
|
|
HLE (with tools)
stem_reasoning
|
73.44%
|
27.08.2026 |
#17/49
better than 67%
|
|
GDPVal-AA v2
general_agent
|
91.48%
|
27.08.2026 |
#9/38
better than 78%
|
|
DeepSWE 1.1
coding_agent
|
76.95%
|
27.08.2026 |
#16/34
better than 55%
|
|
Automation-Bench
general_agent
|
53.85%
|
27.08.2026 |
#9/27
better than 69%
|
What is a good Terminal Bench 2.1 score? What is a good HLE (with tools) score? What is a good GDPVal-AA v2 score?
No comments yet
Be the first to share your thoughts!