Open-weight models
The end of 'tokenmaxxing': enterprises ration AI spend
The brief: buyers are capping AI budgets and routing spend to cheaper Chinese open-weight models, just as OpenAI and Anthropic head for IPOs.
The answer
Enterprises are rationing AI spend as Chinese open-weight models hit ~61% of OpenRouter's top-model traffic.
What happened
Companies including Uber, Microsoft, Salesforce and Meta have moved to ration employees' use of advanced AI, CNBC reported on 26 June. The shift has a name: the end of 'tokenmaxxing' — the practice of rewarding staff for burning as many tokens as possible, with no return-on-investment guardrails. The verdict from buyers is blunt: the token-payment model favoured by Anthropic and OpenAI proved more expensive than it's worth.
Uber is the cautionary tale. CTO Praveen Neppalli Naga said the company blew its entire annual AI budget in four months, then imposed spending tiers starting at $1,500 a month per employee, more on request. Startups are going further. AI-agent firm Lindy shifted 100% of its traffic off Anthropic's Claude to China's DeepSeek; CEO Flo Crivello expects to save millions within months, and says he would switch back if US prices fall.
Crivello described watching the economics flip once he migrated: "you could see that cost curve go down, like, crash to the ground," calling the move to DeepSeek "a matter of survival."
The money
The money is visibly moving. On OpenRouter, a major API-routing platform, Chinese-developed models now account for roughly 61% of token consumption among the top 10 — a near-total reversal in eighteen months.
| OpenRouter top-10 token share | Chinese-developed models |
|---|---|
| Late 2024 | under 1.2% |
| April 2026 | ~51% |
| Late June 2026 | ~61% |
Leaders include Qwen (Alibaba), DeepSeek, Kimi (Moonshot), GLM (Zhipu) and MiniMax. Firms switch for cost and capability — and because open weights let them download a model and run it on their own servers, with no third-party cloud. US startups are now building core products, from code generation to autonomous agents, on these models rather than just demos.
CNBC reports the White House AI crackdown is doing little to slow the trend, with Chinese model makers closing the capability gap even as buyers cite cost as the primary driver of the switch.
What's next
The timing is awkward. OpenAI and Anthropic both filed confidentially for IPOs in early June, and their token-metered business model is now under open scrutiny — Palantir CEO Alex Karp called the token-based approach 'completely wrong' on 1 July. Both labs are already cutting agent costs — Anthropic's Sonnet 5 at an introductory $2/$10, OpenAI's cheaper Terra tier — a direct answer to buyers routing around them.
Frequently asked questions
What does the end of 'tokenmaxxing' mean?
Why are companies switching to Chinese AI models?
How much of OpenRouter's traffic goes to Chinese models?
Does this threaten OpenAI and Anthropic's IPOs?
Sources
- OpenAI and Anthropic face new AI reality as users shift from 'tokenmaxxing' to efficiency — CNBC, 26 June 2026
- White House AI crackdown opens door for Chinese model makers to close gap — CNBC, 30 June 2026
- Palantir's Karp bashes token-based AI model as 'completely wrong' — CNBC, 1 July 2026