Limits & quotas
What counts against your plan, what never does, and the practical limits to keep in mind.
What counts
| Action | Counts as a prompt? |
|---|---|
| New project (Generate) | Yes |
| Fix / revision in the Studio | Yes |
| Refused off-topic prompt | No |
| Failed generation (model busy, bad response) | No |
| Viewing, Live, SEO edits, Git push, import, download | No |
Per plan
| Free | Pro | |
|---|---|---|
| Prompts | 1 in total | Unlimited (fair use) |
| Projects | Created by your prompt | Unlimited |
| Code, zip, Git | No | Yes |
| Generation queue | Standard | Priority |
Practical limits
- One generation at a time per project; start the next fix when the current one finishes.
- Very large prompts are trimmed to what the model can use; put the essentials first.
- Big apps take longer to generate. Split them into a first version and follow-up prompts.
- Live runs in your browser, so very heavy apps may be slower on older devices.
See your numbers anytime on the Usage page in Settings.