deepseektext → text
DeepSeek: DeepSeek V3.1
Canonical ID: deepseek/deepseek-chat-v3.1
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context...
Artificial Analysis Benchmarks
VerifiedNo standardized Artificial Analysis benchmark scores recorded for this variant yet.
Token Pricing & Economics
Prompt Input (1M Tokens)
Input$0.25
Completion Output (1M Tokens)
Output$0.95
Prompt KV-Cache Read (1M Tokens)
Cached$0.13
Context Window Capacity
Context160k tokens
Supported Parameters & Features
frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p