DeepSeek's New AI Model Ranked Cheapest to Run Among Leading Systems -

Chinese artificial intelligence startup DeepSeek has introduced a new AI model that is the least expensive to operate among leading commercial AI systems, according to independent research firm Artificial Analysis, marking another milestone in China’s rapidly advancing AI industry.

The firm’s latest assessment found that DeepSeek V4-Flash offers the lowest inference costs of any well-known AI model while maintaining competitive performance against products from OpenAI, Google, Anthropic and other major developers.

The model reportedly costs US$0.14 per million input tokens and US$0.28 per million output tokens, making it significantly cheaper to run than many of its global rivals.

Artificial Analysis said V4-Flash also performed strongly on its Intelligence Index, matching Google’s Gemini 3.6 Flash and demonstrating capabilities comparable to several premium AI models despite its lower operating costs.

The research highlights the growing emphasis among AI developers on improving efficiency and reducing computing expenses alongside enhancing model performance.

DeepSeek has rapidly emerged as one of China’s leading AI companies, drawing international attention for developing advanced large language models at relatively low cost despite restrictions on access to the most advanced AI chips. The company recently expanded public access to its V4-Flash model and is reportedly working on a more powerful V4-Pro version.

The latest findings come as competition in the AI sector intensifies, with Chinese firms including Alibaba, Moonshot AI, ByteDance and Z.ai accelerating the development of increasingly capable models.

Analysts say the combination of strong performance and lower operating costs is enabling Chinese AI companies to narrow the gap with established Western competitors while making advanced AI services more affordable for businesses and developers.

The report underscores a broader trend in the global AI industry, where reducing deployment costs has become as important as improving model capabilities, particularly as demand for AI-powered applications continues to grow.-ERMD/TS

READ MORE