Deepseek V 3
DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token.
Input
Output
Context
131K
Max Output
8K
Parameters
671B
Technical Specifications
Model TypeChat
Context Window131K
Max Output Tokens8K
Parameters671B
Training Cutoff—
Licensemit
Capabilities
Input Modalities
text
Output Modalities
text
Features
code_generationfunction_callingjson_modereasoningstreamingtext_generation
Benchmarks
Chatbot Arena
#57
Rank (Text)
+4/-4
arena text ci 95
1333
Arena Score (Text)
22,841
Votes (Text)
Artificial Analysis
3.6%
HLE
25.3%
AIME
55.7%
GPQA
35.4%
SciCode
88.7%
MATH 500
75.2%
MMLU Pro
26.0
Math
16.4
Coding
35.9%
LiveCodeBench
16.5
Intelligence
Resources & Links
Related Tools
Data sourced from official provider APIs and documentation
Last updated: May 5, 2026
Start building with the right model.
From model selection to production, one platform, no fragmentation.