---
id: "deepseek/deepseek-v4-pro-0813"
also: ["deepseek/deepseek-v4-pro-20260813", "deepseek-ai/DeepSeek-V4-Pro-0813", "hf/deepseek-ai/DeepSeek-V4-Pro-0813", "requesty/deepseek/deepseek-v4-pro-0813", "vercel/deepseek/deepseek-v4-pro-0813"]
card_usd_per_m: "1.32 in, 3.96 out, at deepseek"
cheapest_list_usd_per_m: "0.0999 in, 8.0 out, at relace/fp4 on openrouter"
cheapest_effective_usd_per_m: "0.0912 in, 3.96 out, at novita on vercel"
context: 1048576
max_output: 943718
input: ["text"]
output: ["text"]
released: "2026-08-12"
reasoning: ["max", "high", "low", "none"]
api: ["/api/v1/chat/completions", "/api/v1/completions", "/api/v1/messages", "/api/v1/responses"]
params: ["frequency_penalty", "include_reasoning", "logit_bias", "logprobs", "max_tokens", "min_p", "presence_penalty", "reasoning", "reasoning_effort", "repetition_penalty", "response_format", "seed", "stop", "structured_outputs", "temperature", "tool_choice", "tools", "top_k", "top_logprobs", "top_p"]
endpoints: 55
gateways: "openrouter (29), vercel (16), hf (5), requesty (5)"
no_training: 40
trains: 6
retains_prompts: 14
intelligence: 23.1000
coding: 68.8000
agentic: 41.3000
---
# DeepSeek: DeepSeek V4 Pro 0813

DeepSeek-V4-Pro-0813 is the official release of DeepSeek-V4-Pro, superseding the preview version, with greatly enhanced agentic capabilities and performance improvements that are especially pronounced in production environments. It is built on the DeepSeek-V4-Pro (Preview) model structure, with a DSpark speculative decoding module attached.