A new generation of cost-efficient Chinese AI models is reshaping the economics of the artificial intelligence industry, forcing American leaders OpenAI and Anthropic to slash prices as they prepare for public market debuts that will demand proof of profitability.
OpenAI cut the price of its GPT-5.6 Luna model by 80% on July 30, reducing API costs from $1 to 20 cents per million input tokens and from $6 to $1.20 per million output tokens, according to CNBC. The company’s mid-tier Terra model also saw a 20% reduction. The moves come as DeepSeek’s V4 Flash — priced at just $0.14 per million input tokens — tops the usage leaderboard on the technical platform OpenRouter.cnbc+2
“I hesitate to say it’s a price war because we’re not quite there yet. But it’s certainly a price competition to try and get more users on board,” technology analyst Jack Gold, founder of J.Gold Associates, told AFP.
Both OpenAI and Anthropic are preparing to go public, which means they “need to start showing a profit,” Gold added. Anthropic filed a confidential S-1 with the SEC on June 1 and could list as soon as October, according to CNBC. The stakes are high: Anthropic’s most recent valuation stood at $965 billion.
Chinese Models Gain Ground
DeepSeek, whose models are open-weight and can be self-hosted by developers, announced last week it plans a “significant increase” in prices for its services — a sign of growing demand, Bloomberg reported. The company currently charges a fraction of what American competitors do: its V4 Flash costs $0.14 per million input tokens compared to $10 for Anthropic’s Fable 5.
New Chinese models from Moonshot AI’s Kimi K3 to Alibaba’s popular Qwen series are grabbing attention as viable alternatives. “Open-source is closing the performance gap with closed models faster than anyone expected,” said Wang Tiezhen, an independent AI consultant and former head of APAC ecosystem at HuggingFace.
Smaller US players are joining the fray. Elon Musk’s xAI has cut prices on some Grok models, and Meta released a low-cost option called Muse Spark 1.2.
The competitive dynamics are already reshaping enterprise AI strategy beyond China and the United States. Indian companies are adopting hybrid approaches, mixing open-weight models for routine tasks with proprietary systems for sensitive work, a shift accelerated by Kimi K3’s mid-July release demonstrating that open-weight models can match premium closed systems.
“AI has become the digital necessity, just like the social media boom some 15 years ago,” John H., CEO of PR News AI said. “While free-tier is good for most of the people, the competition on paid tiers is inevitable, and it’s a great news for the rest of us to be able to pay less for more.”
