Skip to main content
leaderboard
.cn
Toggle Sidebar
Search models or organizations…
⌘K
Explore
Frontier Models
Historical Trends
Local Models
Compare Models
Ranking Studio
Library
Models
Pricing
Organizations
Benchmarks
Data sources
Settings
Toggle Sidebar
leaderboard
.cn
Loading page content
Feedback
Loading page content
Back to models
DeepSeek-V2.5-0905
Open
Share
Add to comparison
Tool use
Overview
Evaluations
Evaluation overview
Overall rank
#151
54.1 points · 180 models
Open-weight rank
#83
58.9 points · 109 models
Data coverage
24 items
1 source · updated through 2024-12-27
Domain performance
Coding
#116 / 160
63.7
Research
#144 / 174
63.9
Long context
#67 / 82
62.6
Evaluation highlights
All evaluations
CLUEWSC
Reasoning · DeepSeek-V3 Technical Report
90.4%
4
/ 9
Aider Edit
Coding · DeepSeek-V3 Technical Report
71.6%
4
/ 7
LongBench v2
Long context · DeepSeek-V3 Technical Report
35.4%
20
/ 22
Arena-Hard
Conversation · DeepSeek-V3 Technical Report
76.2%
20
/ 39
C-Eval
Chinese · DeepSeek-V3 Technical Report
79.5%
35
/ 64
IFEval strict prompt
Instruction following · DeepSeek-V3 Technical Report
80.6%
36
/ 55
Basic information
Context (input / output)
128K / 8K
Parameters
236B / 21B active
Release date
2024-09
License
DeepSeek
Input modalities
Text
Output modalities
Text
Related links
Official site
HuggingFace
Version relationships
DeepSeek-V2 family
DeepSeek-V2-0506
2024-05 · Same series
DeepSeek-V2.5-0905
Current
2024-09 · Current model