
DeepSeek's web chat has never charged individuals directly; the only direct paid channel for individuals has been the API. Its front-end code now contains a complete quota-management system.
IT Home reported on November 14 that developers had found quota-management logic in the chat.deepseek.com front end, including a notice telling users their usage quota is exhausted and to try again once it recovers. As of that report, DeepSeek had not announced pricing, tiers or a launch date for any subscription.
More than an error message
A single new error string could be read as a safety valve for traffic spikes. What the report describes is broader. While the page is visible, the client syncs each model's quota, mute status and server-side copy every 5 minutes. There is a dedicated error branch for an exhausted quota. Even continuing a shared conversation checks an is_exhausted flag in a quotas array. On top of that come reset notifications and server-side usage warnings.
Taken together, this looks built to run for the long haul: quotas synced per model, reset notices, and notice text pushed from the server. Still, a temporary throttle could reuse the same machinery, and the code alone does not rule that out.
The gap between free users and API revenue
IT Home points out that until now the only way DeepSeek charged individuals directly was its API. The API documentation is blunt about it: billing is based on total input and output tokens. The web app is free. There is nothing in between, no way to pay a bit more for a bit more chat.
That gap is particularly relevant to DeepSeek's revenue mix. Figures cited by IT Home—without specifying the reporting period or calculation methodology—put DeepSeek's annualized revenue above $1 billion, almost all of it from the API, while the consumer app still charges users nothing. Last month DeepSeek also merged its Harness team with its Web, App and Platform teams into one product department under a single lead. Against that backdrop, a web quota system reads like groundwork for making the consumer side pay its own way. Per-model limits are exactly what a subscription tier would sell, which is why some community members have tied the code to a plan called "Coding Plan."
It remains an inference, and IT Home labels the claims as community speculation. Metering could just as well be a way to ration free compute more precisely when capacity is tight. Both readings fit, and the code alone cannot tell them apart.
Three things to watch
First, when actual quota numbers and reset windows show up in the interface, since that decides whether ordinary users ever hit the ceiling. Second, whether limits land only on certain models, such as thinking mode or the newest model. Third, whether any paid option appears alongside them. None of these settles the question on its own. Whether this becomes a consumer subscription depends on the formal billing rules: whether users pay on a recurring basis, and what extra quota or features that buys.