Home/DeepSeek/DeepSeek V4
DeepSeek pricing

DeepSeek V4 Cost Calculator

DeepSeek V4 is a 1T-parameter MoE model with frontier-tier reasoning at fraction of US prices. Strong on Chinese and code.

Input
$0.30
per 1M tokens
Output
$1.20
per 1M tokens
Context window
256K
tokens
Released
2026-03
Cutoff 2025-12
≈ Estimated tokenizer·$0.30 in·$1.20 out (per 1M)
Quick start with a use case
Total cost per call$0.000900
Input$0.000300
Output$0.000600
Cost comparison
Standard
$0.000900
With Caching
$0.000770
Save 14% ↓
With Batch
Not supported
Detailed pricing

DeepSeek V4 pricing breakdown

All pricing dimensions including caching and batch discounts.

TypePrice (per 1M tokens)
Input$0.3000
Output$1.2000
Cached input$0.0400

Last verified 2026-05-01 · DeepSeek official pricing

How it compares

DeepSeek V4 vs alternatives

Single-call cost (1000 input + 500 output tokens) ranked from cheapest.

ModelPer call
DeepSeek V4
DeepSeek · this page
$0.000900
GPT-5 mini
OpenAI
$0.000600
Gemini 3.0 Flash
Google
$0.001250
o4-mini
OpenAI
$0.002700
Claude Haiku 4.5
Anthropic
$0.003500
Mistral Large 3
Mistral
$0.006250
Recommended use

When to choose DeepSeek V4

DeepSeek V4 shines for general-purpose tasks, code generation and review, mathematical problem-solving, and complex multi-step reasoning. Token counts are estimated within ~10-20% margin.

Context window of 256K tokens handles long conversations and large documents.
Prompt caching available — significant savings for repeated system prompts.
Tool / function calling supported.
FAQ

Frequently asked questions

DeepSeek V4 costs $0.30 per 1M input tokens and $1.20 per 1M output tokens. A typical chat call (1000 input + 500 output tokens) costs approximately $0.0009. Use the calculator above to estimate your specific use case.