DeepSeek-V3 Ends Promotional Pricing, Updates API Service Rates
Chinese AI company DeepSeek has announced the conclusion of its 45-day promotional pricing period for the DeepSeek-V3 API service, effective February 8. Starting February 9, new standard rates will apply, marking a significant shift from the introductory offers launched on December 26, 2024. Under the updated pricing structure, costs are set at RMB 0.5 ($0.068) per million input tokens for cache hits, RMB 2 ($0.27) for cache misses, and RMB 8 ($1.09) per million output tokens. This represents an increase from the promotional rates of RMB 0.1 and RMB 1 for input tokens, and RMB 2 for output tokens. The price adjustment follows DeepSeek's recent surge in prominence, driven by its model's strong performance in mathematics, coding, and natural language reasoning. Its combination of high capability and competitive pricing has drawn comparisons to OpenAI’s GPT-4, causing notable reactions in Silicon Valley and impacting US tech stocks. This move signals DeepSeek's transition from a market-entry strategy to a sustainable commercial model, reflecting its growing confidence and market position in the global artificial intelligence landscape.
Wire timeline
DeepSeek-V3 Ends Promotional Pricing, Updates API Service Rates
Chinese AI company DeepSeek has announced the conclusion of its 45-day promotional pricing period for the DeepSeek-V3 API service, effective February 8. Starting February 9, new standard rates will apply, marking a significant shift from the introductory offers launched on December 26, 2024. Under the updated pricing structure, costs are set at RMB 0.5 ($0.068) per million input tokens for cache hits, RMB 2 ($0.27) for cache misses, and RMB 8 ($1.09) per million output tokens. This represents an increase from the promotional rates of RMB 0.1 and RMB 1 for input tokens, and RMB 2 for output tokens. The price adjustment follows DeepSeek's recent surge in prominence, driven by its model's strong performance in mathematics, coding, and natural language reasoning. Its combination of high capability and competitive pricing has drawn comparisons to OpenAI’s GPT-4, causing notable reactions in Silicon Valley and impacting US tech stocks. This move signals DeepSeek's transition from a market-entry strategy to a sustainable commercial model, reflecting its growing confidence and market position in the global artificial intelligence landscape.
TechNode