DeepSeek revealed plans on August 6, 2026, to significantly increase prices for its AI API services. The company did not disclose the exact new pricing or when the changes will take effect [1, 2, 3].
Currently, DeepSeek charges US$0.14 per million input tokens and US$0.28 per million output tokens for its V4 Flash model, a rate much lower than rivals in both China and the US. Chinese competitor Moonshot charges US$3 and US$15 per million tokens, while US-based Anthropic's Fable 5 is priced at US$10 and US$50, respectively [1, 2].
DeepSeek released its latest AI model, V4-Flash-0731, on July 31. The model features 284 billion parameters and gained international attention for delivering strong performance at a very low cost [3]. Chinese AI firms have typically relied on aggressive low pricing to compete against US companies, making DeepSeek’s planned hike unusual within China’s AI sector [1, 2].
The company also plans to build a 1 gigawatt AI data center in Inner Mongolia to expand its computing capacity [1].
However, some AI developers expressed concern about the timing of the price increase. Michael Guo, an AI developer, said, "DeepSeek choosing to raise prices at this time – isn’t this just asking for trouble?" He worried the changes could undermine DeepSeek’s competitiveness against new US models like Meta’s Muse Spark and OpenAI’s GPT-5.6 Luna [3].
The coming weeks could reveal more on DeepSeek’s pricing strategy and rollout timeline as competition intensifies in the AI API market.
