Newsletter Subscribe
Enter your email address below and subscribe to our newsletter
[forminator_form id="25163"]

reuters+1techcrunchwtvbam+1The era of "tokenmaxxing" — treating rising AI consumption as a proxy for productivity — is giving way to a more cost-conscious approach, as ballooning bills push enterprises toward open-source and Chinese-built models that deliver comparable performance at a fraction of the price.
Open-source tokens processed on OpenRouter, an AI marketplace, jumped to 65% in June from 34% in January, according to a Citi Citigroup Inc. note cited by Reuters. The shift comes as companies grapple with unpredictable costs driven by AI firms moving from flat subscriptions to usage-based pricing. Uber , for example, burned through its entire 2026 AI budget in just four months after employees rushed to adopt coding tools, forcing management to institute a $1,500 monthly cap per employee, according to TechCrunch and Bloomberg.wtvbam+3
"Changing the license model caught a lot of people by surprise," Harold Byun, CEO of BlueRock, told Reuters. "Immediately after that, we had a number of reports from customers that we're seeing a 20% to 30% spike in terms of over-budgeting". Gartner estimates AI coding costs will surpass the average developer's salary by 2028.reuters+1
The four most popular models on OpenRouter are now all Chinese-made, with DeepSeek holding the top spot. Chinese models charge as little as 18 cents per million tokens compared to an average of $4 for leading proprietary models, according to the Citi note. Top executives including Microsoft's Satya Nadella, Palo Alto Networks' Nikesh Arora, and Coinbase Global's Brian Armstrong have publicly argued that cheaper models can handle a large share of corporate workloads.wtvbam+1
"If you want to win enterprise, you should be forward pricing tokens," Arora wrote on X last week, urging AI labs to charge at the lower rates tokens are expected to command in coming years.wtvbam
The trend puts pressure on OpenAI and Anthropic as both prepare for potential public listings. OpenAI has been reported to be weighing price cuts in anticipation of similar moves from Anthropic. Open-source models are now "90% as good at 10% of the price," said Val Bercovici, chief AI officer at WEKA. The capability gap with frontier models, once estimated at more than a year, has narrowed to roughly four months, according to Byun.instagram+2