OpenAI and Anthropic are expanding cheaper, open-weight Chinese models that are gaining enterprise traction, prompting buyers to weigh price against performance and risk.

As competition intensifies, OpenAI and Anthropic are also lowering prices for their own AI models.

The analysis includes promotional and editorial elements from HokaNews, with disclaimers on financial recommendations, signaling a media-oriented piece rather than a pure corporate press release.

Regulatory and security concerns persist around using Chinese models for private or sensitive data, with ongoing debates over export controls and potential bans.

Chinese rivals DeepSeek and Moonshot are intensifying pressure, shaping pricing strategies and market dynamics.

For enterprises using AI at scale, the model-selection process is shifting from chasing a single top performer to balancing cost, performance, security, and regulatory risk across different workloads.

Revenue and IPO prospects for OpenAI and Anthropic add pressure to stay profitable amid rising compute costs, with Anthropic reportedly profitable in Q2 2026 but not expected to repeat next two quarters; OpenAI aims for a valuation above $1 trillion.

Major US tech firms like DoorDash, Airbnb, and Coinbase are routing workloads to cheaper Chinese models via services such as OpenRouter, signaling growing adoption for non-critical or production workloads.

Price cuts are part of a broader strategy to retain market share amid rising competition from new entrants.

The piece frames a global AI landscape shift driven by price competition among major players.

Chinese model makers such as Moonshot AI, Alibaba, DeepSeek, and z.AI have released benchmark-competitive models, intensifying pressure on US offerings.

Enterprise spend is shifting away from closed proprietary models toward cheaper options, with reduced inference token spend on OpenAI/Anthropic models since mid-July in favor of open-weight alternatives.