What is DeepSeek V4-Pro?

On May 22, 2026, DeepSeek announced that its 75% discount on the V4-Pro model has become permanent. The pricing is now fixed at $0.435 per million input tokens and $1.435 per million output tokens .

Key Features

  • Permanent 75% Discount: Aggressive pricing strategy
  • $0.435/M Input Tokens: Extremely competitive input pricing
  • $1.435/M Output Tokens: Lower than most competitors
  • V4-Pro Architecture: DeepSeek latest model technology
  • API Access: Available via DeepSeek API

Competitive Positioning

DeepSeek pricing is now significantly lower than competitors:

  • Gemini 3.5 Flash: Higher per-token cost
  • Claude 4.7 Opus: Premium pricing
  • GPT-5.5: Premium pricing
  • DeepSeek V4-Pro: Lowest cost in class

Strategic Context

This permanent price drop signals DeepSeek commitment to competing on cost. For developers building agent applications where token volume is high, DeepSeek becomes the economically rational choice .

Pricing

  • Input Tokens: $0.435 per million
  • Output Tokens: $1.435 per million

Pros

  • Lowest cost among major LLM providers
  • Permanent pricing — no expiration
  • Ideal for high-volume agent applications
  • Competitive performance with V4-Pro
  • API accessible

Cons

  • China-based provider (data governance considerations)
  • May have different content policies
  • Less brand recognition than US providers
  • Documentation primarily Chinese

Who Should Use It?

Perfect for: Cost-conscious developers, high-volume agent applications, startups with tight budgets, and teams building token-intensive applications.

Verdict

DeepSeek permanent 75% discount changes the LLM economics for high-volume applications. At $0.435 per million input tokens, it the most cost-effective option available .

Rating: 4.4/5 - The cost king of LLMs.