Hub/Models/deepseek/deepseek-v4-pro
deepseektext → text

DeepSeek: DeepSeek V4 Pro 0423

Canonical ID: deepseek/deepseek-v4-pro

Compare ModelOpen on OpenRouter →

DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reasoning, coding,...

Artificial Analysis Benchmarks

Verified
Intelligence Index20.8 / 100
Engineering Index28.7 / 100
Strategy & Ops36.6 / 100
Finance & Accounting28 / 100
Legal Index26.3 / 100

Token Pricing & Economics

Prompt Input (1M Tokens)
$0.90
Input
Completion Output (1M Tokens)
$1.50
Output
Prompt KV-Cache Read (1M Tokens)
$0.20
Cached
Context Window Capacity
1,024k tokens
Context

Supported Parameters & Features

frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_completion_tokensmax_tokensmin_ppresence_penaltyreasoningreasoning_effortrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p