DeepSeek Slashes AI Costs to Cents
DeepSeek has permanently reduced the price of its V4 Pro AI model by 75%, making it significantly cheaper than competitors. The new pricing structure allows enterprises to save substantial amounts on AI workloads, particularly for high-volume tasks. This change is attributed to the model's architectural efficiency and innovative Mixture-of-Experts design.
- ▪DeepSeek's V4 Pro model now costs $0.87 per million output tokens, down from $3.48.
- ▪The model's Mixture-of-Experts architecture activates only a fraction of its parameters, reducing costs while maintaining performance.
- ▪Switching from GPT-5.5 to V4 Pro can save enterprises approximately $2.4 million per year.
Opening excerpt (first ~120 words) tap to expand
DeepSeek Slashes AI Costs to CentsEdition #299 | 29 May 2026Business Analytics NewsletterMay 29, 202681ShareDeepSeek Makes 75% Price Cut on V4 Pro Permanent, Dropping Frontier-Class Inference to $0.87/M Output Tokens with Mixture-of-Experts ArchitectureIn this edition, we will also be covering:Insilico teams up with US firm to build AI models predicting disease decades aheadChina unveils auto industry blueprint to set EV, AI vehicle and semiconductor standardsOpenAI's Altman says AI unlikely to lead to 'jobs apocalypse’Today’s Quick WinsWhat happened: DeepSeek has made permanent the 75% price reduction on its flagship V4 Pro model, locking in rates of $0.435/M input and $0.87/M output tokens down from the previous $1.74/$3.48 per million tokens.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Hacker News (AI / LLM).
Discussion
0 commentsMore from Hacker News (AI / LLM)
