DeepSeek Slashes API Prices by 75% for V4-Pro Until May 5th, 2026
Cheapest frontier model yet? DeepSeek V4-Pro costs 75% less for devs.
On April 27, 2026, Chinese AI startup DeepSeek announced aggressive price reductions for its newly unveiled DeepSeek V4-Pro model, offering a 75% discount on API usage to developers until May 5, 2026. This dramatic cut makes V4-Pro one of the most affordable high-performance frontier models available, potentially undercutting rivals like OpenAI’s GPT-4o and Anthropic’s Claude Opus. Additionally, DeepSeek slashed prices for input cache hits across its entire API lineup to just one-tenth of the original cost, a move that significantly reduces latency costs for applications with repeated queries.
These pricing changes are clearly designed to accelerate developer adoption and intensify competition in the increasingly crowded AI market. By offering such steep discounts for a limited time, DeepSeek aims to lure developers away from established providers and build a loyal user base for its ecosystem. The cache hit reduction is particularly strategic—it lowers the total cost of ownership for production deployments that rely on caching patterns, making DeepSeek’s API stack more attractive for high-volume, real-time applications like chatbots, coding assistants, and retrieval-augmented generation (RAG) pipelines.
- DeepSeek V4-Pro API gets a 75% price cut until May 5, 2026.
- Input cache hit prices reduced to one-tenth the original across all API models.
- Move aims to boost developer adoption and compete with OpenAI and Anthropic.
Why It Matters
DeepSeek's aggressive pricing forces competitors to match, lowering AI inference costs for every developer.