DeepSeek Instead of OpenAI: US Companies Switch to Cheaper AI

According to CNBC, in a report published in July 2026, US companies are increasingly switching to Chinese AI models, primarily DeepSeek, because of the fast-rising costs of OpenAI and Anthropic solutions.

Why US Companies Are Turning to Chinese Providers

The main reason is the price gap. Using the OpenAI API or Anthropic's Claude costs organizations significantly more than comparable services built on Chinese models. Meanwhile, the quality of those models, which until recently trailed the market leaders noticeably, keeps closing in on the American flagships.

Small and medium-sized businesses that need bulk data processing, content generation or customer support automation proved especially sensitive to API request costs. This segment started switching first, particularly companies that use AI not as an experiment but as operational infrastructure.

An important nuance: this is not a complete abandonment of American providers. Many companies are moving to a mixed model. Tasks that demand top quality or involve sensitive data stay with OpenAI or Anthropic, while high-volume routine operations go to cheaper Chinese models.

What Has Changed in the Quality of Chinese Models

As recently as 2024, Chinese AI models performed noticeably worse than GPT-4 and Claude 3 on complex tasks such as multi-step reasoning, programming and analysis of long documents. By mid-2026 the picture had changed: a number of Chinese developers, including the teams at DeepSeek, Alibaba and ByteDance, released models whose results on standard benchmarks are comparable to the leading Western flagships.

For most practical business tasks, such as document summarization, customer chatbots, handling incoming requests and basic coding, the quality gap between the best American and Chinese models has shrunk to a negligible level. That is what makes the price argument especially persuasive.

• Chinese models cost several times less than American ones in API token pricing
• Quality on practical tasks (documents, support, code generation) has converged substantially with OpenAI and Anthropic
• Companies with high request volumes are switching first: retail, service platforms, SaaS
• Some companies use a hybrid approach: expensive models for complex tasks, cheap ones for routine work

How OpenAI and Anthropic Are Responding to the Competition

American AI labs are building differentiation on more than model quality. Their key arguments are data privacy and compliance with US regulatory requirements, which matter fundamentally to corporate clients in finance, healthcare and defense.

OpenAI and Anthropic are also investing heavily in deep integration with enterprise systems, developer tools and partner ecosystems, areas where open or Chinese models still lag in maturity. Even so, price pressure is growing: since 2023, token prices at the leading American providers have fallen by dozens of times, and the trend continues.

What This Means

The shift of US companies to Chinese AI models signals structural maturity in the market: when baseline quality converges across providers, price becomes the main deciding factor. For OpenAI and Anthropic, it is a signal to speed up differentiation through enterprise tooling, data security and industry-specific solutions. For the market as a whole, it means faster commoditization of basic AI services and further price cuts across the sector. Companies that were overpaying for AI infrastructure two years ago now get comparable quality for less money, and they are taking advantage of it.

Why are companies switching to DeepSeek?

The main reason is the price gap. Using the OpenAI API or Anthropic's Claude costs organizations significantly more than comparable services built on Chinese models.