Bolt pricing: monthly Pro, tokens and publication capacity

Use a directly verified monthly offer

The checked public monthly offer lists Pro at USD25 and Teams at USD30 per member, with Pro starting at ten million monthly tokens. Applicable taxes are additional. An earlier annual-equivalent figure is a different billing arrangement and should not be displayed as the monthly checkout amount. Inspect the selected capacity, final total and billing date when upgrading. The useful entry-price question is whether the capacity required by your project is available under the period you intend to purchase, rather than whether an older review contains a smaller monthly-looking number.

Check daily and monthly Free limits

Free documentation describes a daily three-hundred-thousand and monthly one-million token limit. A daily reset does not remove the monthly boundary. One successful free request is therefore different from enough capacity for a week of revisions. Test the first build, one correction and one new requirement, and record where work stops. If repeated attempts come from ambiguous requirements, improve those before purchasing more capacity. The trial should identify whether the bottleneck is the allowance or the way the task is described, instead of treating every interruption as evidence that a paid tier will solve it.

• Choosing Bolt: tokens, code review and publishing

Measure more than generated text

Tokens are used for project-reading work as well as visible output. A one-character revision in a large codebase need not consume resources in proportion to the short instruction. Keep relevant types, callers and reproduction evidence while removing unnecessary scope. Record investigation and retries through the accepted change. Do not divide another vendor’s credits or message allowance by Bolt tokens as though the units were interchangeable. Compare equivalent finished work and human checking time. That gives a basis for selecting capacity rather than assuming the largest printed token number is automatically the best value.

Hypothetical capacity planning

For a hypothetical capacity plan, use C × N + R: your measured average small-change consumption, planned count and a reserve for larger investigation. C is not an official guaranteed task price. Inspect outliers and rejected attempts too. Reload eligibility depends on the plan and billing arrangement, so do not assume identical additional-token purchasing is available everywhere after exhaustion. Compare a higher allocation, an eligible reload and waiting for refresh using the actual workload schedule. The right choice follows from the interruption you need to avoid, not simply from buying the largest allowance in advance.

• Lovable pricing and credits: shared balance, limits and cancellation

Review hosting and collaboration separately

Free publication has hosting and domain conditions different from paid access. Billing documentation describes Free hosting limits and possible downtime after exceeding them. Remaining development tokens alone do not establish operating continuity. For Teams, distinguish headcount from individual consumption and inspect the documented member allocation rather than assuming a pooled token wallet. Inventory hosting, required databases and user growth separately from the builder’s work. A pricing estimate should support both the developers making changes and the people using the published application, without letting one allowance stand in for the other.

Separate cancellation and preservation

Official cancellation guidance keeps the paid plan through the current period and moves to Free afterward. Inspect remaining-token and rollover conditions instead of assuming indefinite use of the same balance. Preserve GitHub source, required data, configuration names and publication ownership. Verify publication separately if the app should stop. Have the next maintainer complete a small revision in the intended destination before retiring connections. Record cancellation and asset preservation as distinct completed outcomes. This prevents a financial change from being mistaken for a complete operating migration or a source backup from being mistaken for ending renewal.

• Checking accessibility of AI-generated UIs before release