Gemini just copied Claude's pricing model. That makes ChatGPT the best value in compute. It might also prevent Anthropic from raising their price.

Google switched Gemini from a fixed number of daily requests to “compute-based” usage limits. Prompt complexity, features used, chat length. It all counts now. GitHub did the same with Copilot last month, moving to token-based “AI Credits.”

Anthropic has been ahead of this curve. Claude’s Pro and Max plans already run on token consumption under the hood. They added an “extra usage” toggle that lets you keep working past your limit at API rates instead of hitting a wall. And they have been transparent that their flat rate plans were never built for agentic workloads like Claude Code.

Thanks to Google, this move actually helps out the Claude user.

I’ve been saying for a while that Claude Cowork and Code are WAY too popular at the current price point.

It’s too good of a deal for the productivity and value. We were due for a price increase from Claude on their $100/month plan.

But, instead of Anthropic raising the price of tokens, they came out with a better model that uses tokens faster. So, we essentially got the price increase. However, this move by Google helps keep them in-check since their token usage model is not published (yet). They will keep Anthropic guessing and probably stalled raising the price for a minute.

OpenAI is the holdout. ChatGPT still uses message counts on rolling 3 hour windows. 160 messages every 3 hours on Plus. Simple, but it treats a 2 sentence question the same as a 10 turn agentic research task. That math will break eventually.

Three different approaches. Google went vague (“compute-based” with no published formula). Anthropic went transparent (tokens, with overflow pricing you control). OpenAI is pretending the problem does not exist yet.

Let the battle of tokenized billing begin! It can only be a race to the bottom as the cost of compute gets cheaper.

Are we having fun yet?

https://www.pcworld.com/article/3142744/google-just-made-big-changes-to-gemini-usage-limits.html