deepseek

Deepseek V3

text-to-text
Efficient Mixture-of-Experts model with 37B active parameters per token, excelling in coding, math, and reasoning tasks while maintaining 128K token context length.
Model Parameters

Advanced settings to control the behavior of the model.

A system prompt that will be prepended to all chat messages to guide the model's behavior. Automatically gets saved to your playground repository.

Controls randomness: 0 = deterministic, 2 = very creative

Maximum length of the response (1-128000 tokens)

deepseekDeepseek V3

Log in to chat with any model.

Connecting to Model...