DeepSeek tells clients to expect a significant API price hike
DeepSeek has begun notifying its customers of an unspecified but significant increase to its API rates, putting the China-based AI provider’s reputation as the most affordable option in the market on a countdown. The company has not disclosed how much prices will rise or when the new rates will take effect, with the notice apparently intended to give clients time to adjust.
The coming change is separate from a mid-July adjustment in which DeepSeek raised rates during peak hours to manage server load, effectively doubling API usage prices in certain time windows.
What DeepSeek currently charges
DeepSeek’s V4 Flash model is priced at $0.14 per million input tokens and $0.28 per million output tokens. By comparison, Google’s Gemini 2.5 Flash-Lite, marketed as a cost-efficient option, is priced at $0.10 per million input tokens and $0.40 per million output tokens. The actual cost to any given user depends on the mix of input and output tokens their workloads consume.
Why the change matters
DeepSeek has built a large share of its customer base on price competitiveness against providers such as OpenAI, Google, and Anthropic. A significant rate increase would narrow or remove that gap, particularly for workloads that lean heavily on output tokens, where DeepSeek is currently cheaper than Gemini 2.5 Flash-Lite.
DeepSeek has also drawn positive attention for the capability of its V4 Flash model, which launched ahead of the pricing notice and gave the company momentum that the upcoming change could partially offset.
Possible funding for a new data center
Separately, DeepSeek has outlined plans to construct a large-scale data center in Inner Mongolia to expand its AI infrastructure, though the company has not officially confirmed the project. Higher API rates could serve as one of several revenue sources to fund the buildout, alongside the existing peak-hour surcharges.
What customers should watch
Clients waiting on firmer numbers should monitor DeepSeek’s official channels for an updated rate card and an effective date. Teams with token-heavy workloads may want to benchmark current spend against Gemini 2.5 Flash-Lite and other low-cost alternatives now, so the impact of any new pricing can be modeled before invoices change.
FAQ
How much are DeepSeek’s API prices going up?
DeepSeek has not said. The notices describe the increase as significant but do not include a specific figure or an effective date.
Is this the same as the July peak-hour price hike?
No. The July change raised API rates during certain peak hours to reduce server load. The new notice is a separate, broader increase to standard rates.
Why might DeepSeek be raising prices?
DeepSeek has set out plans for a large data center in Inner Mongolia to expand its AI infrastructure. Higher API revenue could help fund the buildout, though the company has not officially confirmed the project or linked it directly to the pricing change.
This article summarizes reporting from androidauthority.com.