DeepSeek's V4-Flash AI Model Identified as Cheapest to Run Among Well-Known Models

A new report by Artificial Analysis indicates that DeepSeek's V4-Flash is the most cost-effective well-known AI model to operate, with benchmark testing costing approximately three cents. This significantly undercuts rivals like Moonshot's Kimi K3 (86 cents), OpenAI's GPT-5.6 Sol ($1.86), and Anthropic's Claude Fable 5 ($3.15). DeepSeek's published pricing is $0.14 per million input tokens and $0.28 per million output tokens, positioning V4-Flash as a high-volume, fast, and affordable option, though its capability score on the Intelligence Index is 50 out of 100, comparable to Google's Gemini 3.6 Flash.