US AI Labs Cut Prices as Chinese Competitors Win Budget-Conscious Clients

OpenAI and Anthropic are reducing the cost of their models in response to a growing number of businesses switching to cheaper alternatives from Chinese developers. The price competition follows rising AI expenses that have pushed companies to limit usage and explore more affordable options, with firms like Moonshot and DeepSeek gaining traction among users in the US and Europe.

OpenAI recently announced an 80 percent price cut for GPT-5.6 Luna, which it describes as its fastest and most affordable model. Anthropic introduced Claude Opus 5, claiming it offers "frontier intelligence" at half the price of its top-tier Fable 5. According to Silicon Data's token price index, these moves have reduced what customers pay for leading US lab models by nearly a quarter since mid-July. Tokens are the units of data processed by language models and often determine customer bills.

The price cuts mark a shift for US firms that have traditionally competed on performance with proprietary "closed" models. Freely downloadable and modifiable "open" Chinese models have added to the pressure. The changes come as both companies consider initial public offerings at trillion-dollar valuations, with investors looking for proof that their massive AI investments can yield returns.

Corporate clients are feeling the pinch as Anthropic and OpenAI move some enterprise customers from flat subscriptions to usage-based billing, where charges depend on computational resources used. In response, some businesses have set caps on AI usage or tested cheaper options. DoorDash and Airbnb have confirmed they now use Chinese-made models to lower their bills. Meanwhile, Chinese labs have released a wave of models that have narrowed the performance gap with leading US systems, raising concerns that American developers could lose customers despite heavy spending to stay ahead.

Source: Ars Technica