Urgent.News

What's breaking now, across thousands of outlets.

AI

DeepSeek Raises Prices by Up to 4x

DeepSeek built much of its reputation by offering powerful AI models at significantly lower prices than many Western rivals. That … Read More The post DeepSeek Raises Prices by Up to 4x appeared first on ProPakistani .

DeepSeek Raises Prices by Up to 4x

DeepSeek, a Chinese AI company, announced today that it will be significantly increasing the prices of its V4 models by up to four times. The price hike comes into effect on August 16, and the changes are part of the company's new peak and off-peak pricing system designed to optimize computing resource allocation. DeepSeek V4 Pro's price, which was originally set at $0.87 per million output tokens, will skyrocket to $3.96 during peak hours, a more than fourfold increase.

Off-peak pricing, which occurs between 11:00 PM and 3:00 AM, will see a less severe price rise to $1.98 per million output tokens. DeepSeek's cheaper V4 Flash model will also see an increase, with its peak-hour rate rising from $0.28 to $1.32 per million output tokens and its off-peak rate from $0.14 to $0.66 per million output tokens.

Despite the price increase, DeepSeek's V4 Pro remains cheaper than several competing flagship AI models, such as Moonshot AI's Kimi K3, which costs around $15 per million output tokens, and OpenAI's GPT-5.6 Sol, priced at $30 per million output tokens. However, OpenAI's lower-cost GPT-5.6 Luna, priced at $1.20 per million output tokens, is slightly more affordable than DeepSeek's V4 Flash during peak hours.

Written by urgent.news from ProPakistani's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at propakistani.pk →

More in AI

Private AI Inference with Homomorphic Encryption: A Practical Guide to Computing on Encrypted Data

In 2009, Craig Gentry proved that it is possible to compute on encrypted data without ever decrypting it, and the result was widely treated as a theoretical curiosity.

  • Homomorphic encryption enables computations on encrypted data without decryption.
  • CKKS scheme supports machine learning computations on encrypted data.
  • Private AI inference pipelines face cost considerations for feasibility.

More from Saturday 15 August →