DeepSeek Cuts V4 Pro Prices 75%, Igniting a Chinese AI Pricing War

DeepSeek aggressively cut the price of its flagship V4 Pro model to $0.0036 per 1M input tokens and $0.87 per 1M output tokens — a 75% reduction from launch pricing — repositioning the model around 'AI performance efficiency per dollar.' The move triggered a surge in demand and forced rival Chinese AI firms and cloud providers to recalibrate their own pricing, per Caixin and SCMP reporting.

The cuts arrive as DeepSeek prepares a new funding round, and developers read the combination as a strategic shift: China moving from 'model release' to 'capital + infrastructure + deployment.' By undercutting on cost while a raise looms, DeepSeek is buying market share and developer mindshare simultaneously.

The pricing war has ripple effects beyond China. Community discussion noted OpenAI is weighing token price cuts to counter competitive pressure, with predictions that token prices will continue to plummet industry-wide as labs bleed billions on compute.

Watch whether Western providers respond with matching cuts, how margins hold up across the sector, and the terms of DeepSeek's pending round.