Model Terminal

DeepSeek V3 0324

DeepSeek V3 0324 is a large Mixture-of-Experts (MoE) language model with 671B total parameters (37B active per token), released by DeepSeek in March 2025. It targets general-purpose chat, code generation, long-context reasoning, and tool use, and is published as an open-weight model. Source: supabase.

Value score
Not scored
Context
tokens
Max output
tokens
Price

Capability radar

Identity

Developer
~deepseek
Openness
Modalities
Release
Knowledge cutoff
Deprecation
API docs

Benchmarks

No benchmark scores yet.

Compare nearby

Pricing

Input / 1M
Output / 1M
Speed