Max Output Tokens
The ceiling on how many tokens a model will generate in a single response — separate from the context window.
Max output tokens is the hard cap on the length of a single response. It is distinct from the context window: the window is the total budget for prompt plus completion, while the output cap limits only the completion. A model with a 1M-token window might still cap a single answer at 128K output tokens, so a task that needs an enormous response can hit the output ceiling long before it exhausts the window.
This matters for long-form generation — big code files, exhaustive reports, bulk translations — where you may need to continue across multiple calls to get the full result, paying output rates each time. It also matters for reasoning models, whose internal reasoning consumes the output budget alongside the visible answer.
This explorer lists the maximum output where the provider publishes a distinct figure; many providers do not break it out separately on the pricing page, in which case it shows a dash rather than a guess. When you are planning a workload that produces very long outputs, confirm the current cap in the provider’s API docs, since these limits change more often than prices do.
FAQ
What happens if my response hits the max output limit?
The model stops generating at the cap. To get a longer result you continue in a follow-up request, paying output rates for the additional tokens.
Why do some models show no max output figure?
Because the provider does not publish a distinct output cap on its pricing page. This explorer shows a dash rather than inventing a number.