Select Language:
Thanks to tips from several IT之家 readers, including 软媒新友2108385, 锐意行省, 南琴梨, ZERO_A_ONE, SJJ, 软媒用户1812279, and GreatMOLA, we have some important updates regarding DeepSeek’s API pricing.
On May 22, IT之家 reported that DeepSeek’s official announcement states a significant change for their DeepSeek-V4-Pro model API. After the conclusion of their promotional 75% discount that ends on May 31, the API’s prices will be permanently reduced to just a quarter of their original cost. This effectively maintains the 75% discount indefinitely, rather than reverting to the standard pricing in June as initially planned.
The API’s previous rates were as follows:
– For cached inputs: 0.1 yuan per million tokens
– For uncached inputs: 12 yuan per million tokens
– For outputs: 24 yuan per million tokens
The company’s latest notice emphasizes that this price cut is not temporary — the API pricing will stay at these reduced levels permanently, making it a unique opportunity for developers and businesses to access large language model services at a significantly lower cost.
Visuals accompanying the announcement highlight the stark contrast between current and future pricing, reinforcing the size of this ongoing discount. This move points to DeepSeek’s commitment to providing more affordable AI solutions and could influence pricing strategies across the industry moving forward.
ChatGPT
Perplexity
Gemini AI
Grok AI
