DeepSeek announced on Thursday that it plans to raise prices across its API services, with the compa...
The AMW Read
Price increase signals a strategic pivot for DeepSeek within the foundation model segment, updating its positioning as a cost leader; cross.§B is relevant as the article explicitly ties pricing to model release and compute demand.
DeepSeek announced on Thursday that it plans to raise prices across its API services, with the company saying a relatively large increase is expected and that it will announce the final pricing scheme separately. The move comes less than a week after DeepSeek began public testing of the official API version of V4-Flash on July 31, and industry experts suggest the price hike also signals that the official version of V4-Pro is close to launch, which would increase overall computing demand.
This marks a notable shift for DeepSeek, whose aggressively priced models have helped drive down the cost of deploying AI applications and intensified competition among Chinese model developers. The company currently uses peak and off-peak pricing, with peak periods (9 am to noon and 2 pm to 6 pm on weekdays in Beijing) at double the standard rate. Under regular pricing, V4-Flash costs RMB 1 ($0.14) per million input tokens (cache miss) and RMB 2 per million output tokens, while V4-Pro costs RMB 3 per million uncached input tokens and RMB 6 per million output tokens.
The price increase could reshape the competitive dynamics in China's foundation model market, where DeepSeek has been a key price disruptor. For builders and enterprises relying on DeepSeek's API, the change is a prompt to reassess cost structures and explore alternative providers or negotiation strategies. For investors, it signals that DeepSeek, despite its history of aggressive pricing, is moving toward a more sustainable revenue model, potentially ahead of its planned IPO and external fundraising rounds. #DeepSeek #AI #pricing #ChinaAI #APIs #foundationmodels
