LLM Reference

DeepSeek V3 on Together AI

DeepSeek V3 · DeepSeek

Serverless

Capabilities

VisionMultimodalReasoningFunction CallingTool UseJSON ModeCode Execution

About DeepSeek V3

DeepSeek's V3 model with 671B total parameters (37B active via MoE). State-of-the-art performance on reasoning and coding benchmarks.

Get Started

Model Specs

Released2024-12-26
Parameters671000000000
Context64k
ArchitectureMixture of Experts
Knowledge cutoff2024-04