All prices › DeepSeek › DeepSeek V4 Pro

DeepSeek V4 Pro API pricing

As of October 16, 2026, DeepSeek V4 Pro costs $1.32 per million input tokens and $3.96 per million output tokens on the DeepSeek API. Cached input is $0.044 per million.

It is the most expensive of 7 DeepSeek models tracked here, by blended price. The closest Standard-tier model from another provider is Claude Haiku 4.5 (Anthropic) at $1.00 / $5.00; DeepSeek V4 Pro is 1% cheaper. First listed in the LiteLLM price data on June 16, 2026. Its price has been cut 0 times and raised 1 time since.

Input / 1M tokens$1.32
Output / 1M tokens$3.96
Context window1M
Max output393K

Every published rate

RatePer 1M tokens
Input (standard)$1.32
Output (standard)$3.96
Cached input (read)$0.044

What it costs in practice

TaskTokensPer requestPer 1,000 requests
Chatbot reply1,000 in / 500 out$0.0033$3.30
Summarize a long document20,000 in / 1,000 out$0.03$30.36
Coding agent step50,000 in / 2,000 out$0.07$73.92

Standard rates, no caching or batch discount. Compare your own numbers across all models →

Price history

$3.96 $0 2026-06-162026-10-16
Per 1M tokens. ━ Output · ┅ Input
DateModelChange
2026-08-19DeepSeek V4 ProOutput price: $0.87 → $3.96 (+355%) source
2026-08-19DeepSeek V4 ProInput price: $0.435 → $1.32 (+203%) source
2026-06-16DeepSeek V4 ProNew model listed: $0.435 in / $0.87 out source

Dates are when each change first appeared in the LiteLLM price data, usually within days of the provider's announcement.

Details

ProviderDeepSeek
Price tierStandard
FeaturesReasoning, Prompt caching, Tool use
API model namesdeepseek-v4-pro
Official pricingapi-docs.deepseek.com

Models at a similar price

ModelInput / 1MOutput / 1M
Grok 4.20$1.25$2.50
Grok 4.3$1.25$2.50
GPT-5.4 Mini$0.75$4.50
o3 Mini$1.10$4.40
o4 Mini$1.10$4.40
Claude Haiku 4.5$1.00$5.00

Other DeepSeek models

ModelInput / 1MOutput / 1M
DeepSeek R1$0.55$2.19
DeepSeek V4 Flash$0.30$1.20
DeepSeek V4 Flash Vision Exp$0.30$1.20
DeepSeek V3$0.27$1.10
DeepSeek V3.2$0.28$0.40
DeepSeek Coder$0.14$0.28