According to the AI support bot that running in max context mode, even when under the max context window (E.g. the first 200k) the token cost is 1.2x instead of 1.0. Shouldnt the “max” mode not incur any charges until you exceed the normal token amount? seems very strange there would be an extra cost for the same thing unused?
Steps to Reproduce
Run max mode in a new window, see higher per-token costs.
Expected Behavior
Costs are identical between max and non-max until exceeding the 200k base token window (Anthropic for example)
Hey, this is expected behavior. Max Mode always applies a 1.2x multiplier to the API cost, no matter how much context is actually used in the conversation.
Max Mode consumes usage at 1.2x the normal API rate for the selected model.
The 1.2x is the fee for access to Max Mode features like extended context, subagents with models other than Composer-1, and image generation on request-based plans, not for actually going over the base context window.
Hi Dean, thats really not very transparent, it should be called out in the text/on the pricing page.
Also, it seems pretty crazy. Happy to pay for extended context, when using it, but not int he run up. Claude does not charge for this, neither should Cursor. Actually it’s a double hit because you charge 20% for max mode and another 20% on top of that for the token fee or whatever it is. If you can share disgruntlement with your product team it would be appreciated.
Also, gotta rename one of the max’s. Its confusing as hell. Max thinking, max context. Pick max for one, use extended for the other or something.
Max Mode extends the context window to the maximum a model supports. More context gives models deeper understanding of your codebase, leading to better results on complex tasks. The models table above shows each model’s maximum context size.
Max Mode uses token-based pricing at the model’s API rate, so it consumes usage faster than the default context window. On individual plans, a 20% upcharge is added to the model’s API rate. On Teams and Enterprise plans, the Cursor Token Fee applies instead of the 20% upcharge.
—
It’s not accurate since the model API rate is not increased until you exceed 200k tokens. There is a hidden 20% here. It should be just the token rate on top of the base model rate. But thats not what appears to be happening.
e.g. from the usage dashboard there are 4 variants:
I Agree with @charles . Max Mode is not transparent enough and the upcharge of 20% for using Max Mode even when not going above the normal context window size (e.g. 200K) is ridiculous. Max Mode should be integrated automatically and trigger when the context window gets too big. That makes sense and would be fair.
Sadly after spending tens of thousands on Cursor, I’ve had to abandon it for claude code. Results are typically better, and actual spend cost is 1/4. Sorry folks, great product, but not manageable pricing models. At the end of the day, there’s not enough value add over the other tools to warrant using Cursor anymore.
