DeepSeek V4-Flash: Cheapest AI Model to Run, Study Says

A version of Chinese startup DeepSeek's flagship AI model is the least expensive to run among well-known models globally. This is according to a research firm. Here's the thing: DeepSeek's V4-Flash model costs 100 times less to run than Anthropic's Claude Fable 5. The company officially released this V4-Flash model on Friday. DeepSeek is known for offering ultra-low-cost AI alternatives. The V4-Flash charges 14 cents per million input tokens and 28 cents per million output tokens. A research firm estimates its average cost at three cents per test. This compares to 86 cents for a rival Chinese model and over three dollars for Claude Fable 5. What's interesting is this comparison provides a realistic measure of value. It accounts for the data a model processes and generates to complete a task. The V4-Flash model scored 50 out of 100 on an Intelligence Index. This is the same score as Google's Gemini 3.6 Flash. This cost-effectiveness could be a game-changer for businesses looking to deploy AI at scale.