TLDR
- OpenAI cut the price of its GPT-5.6 Luna model by 80% and trimmed Terra pricing by 20%
- Anthropic launched Claude Opus 5 at half the price of its previous top model, Fable 5
- Anthropic also cancelled a planned price hike for Claude Sonnet 5
- Companies like DoorDash, Airbnb, and Siemens are testing cheaper Chinese AI models from DeepSeek and Moonshot AI
- Chinese models have overtaken Claude and ChatGPT in token use on the OpenRouter platform
OpenAI and Anthropic are cutting prices on several AI models after enterprise customers began shifting toward cheaper Chinese alternatives like DeepSeek and Moonshot AI.
OPENAI AND ANTHROPIC IN PRICE WAR AS CHINESE AI RIVALS GAIN GROUND-FT
— First Squawk (@FirstSquawk) August 14, 2026
Why Prices Are Falling
Businesses that once spent freely on AI are now tracking costs more closely. Chinese AI developers have stepped in with capable models at a fraction of the price, giving companies real alternatives for everyday tasks like coding and customer support.
The numbers tell the story. One million output tokens on Anthropic’s Fable model costs $50. The same volume on Moonshot AI’s Kimi K3 costs $15. On DeepSeek V4 Pro, it costs just $0.87. That gap is hard for enterprise buyers to ignore.
OpenAI responded by cutting the price of its GPT-5.6 Luna model by 80%, dropping from $1 to $0.20 per million input tokens. Its mid-tier Terra model was trimmed by 20%. The flagship Sol model kept its price but got a faster processing mode.
Anthropic moved too. The company launched Claude Opus 5 at $5 per million input tokens and $25 per million output tokens, positioning it at half the cost of its former top model, Fable 5. It also cancelled a planned September price increase for Sonnet 5, keeping the lower introductory rate in place.
Which Companies Are Switching
The shift to Chinese models is already happening at major firms. DoorDash has been using Moonshot AI for coding tasks. Airbnb and Siemens are testing models from DeepSeek and Alibaba.
Data from OpenRouter shows Chinese models from DeepSeek and Z.ai have overtaken both Claude and ChatGPT in token usage on the platform. DeepSeek held the top spot for model usage by token share during that period.
That said, cheaper tokens do not always mean lower total costs. A more capable model may finish a task using fewer tokens and with fewer errors. Experts say the better measure is cost per successful task, not price per million tokens.
Even when Chinese models do not fully replace US systems, their presence gives enterprise buyers more leverage in pricing negotiations.
The pressure on OpenAI and Anthropic is real. Both companies face high infrastructure costs to build and run advanced models, even as they lower prices to keep customers. The question now is whether these cuts can hold as AI usage scales up.
For businesses, the trend brings more choice and lower costs. For the AI companies themselves, it means the race is no longer just about building the most powerful model. Price is now a core part of the competition.
Stop guessing and start investing with confidence. KnockoutStocks gives you the AI insights, market intelligence, and stock research you need to spot opportunities, cut through the noise, and make smarter investment decisions — all in one powerful platform.
Sign up today and get 50% OFF full access to our premium stock picks.
Simply use coupon code SPECIAL50 at checkout to claim your exclusive discount.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
