Skip to main content

Quotas & Limits

PrompTick measures usage in two user-facing units: AI Actions and Test Runs. This page explains exactly what counts.

AI Actions​

An AI Action is one completed AI workflow initiated by you. Each of these counts as one AI Action:

  • Generating a new prompt (generate_prompt)
  • Improving an existing prompt (improve_prompt)
  • Enriching a coding prompt (enrich_coding_prompt)
  • An execution of one of your deployed agents triggered from the PrompTick UI
  • An MCP tool call from an AI coding assistant (Claude Code, Cursor, etc.)

What does NOT count as a separate AI Action​

  • Internal sub-agent calls. A single workflow (e.g. generate_prompt) may internally orchestrate multiple AI steps. The whole workflow counts as one AI Action regardless of how many sub-steps run under the hood.
  • Clarification round-trips within the same workflow. If a workflow asks you a clarification question and you answer it, the follow-up is part of the same AI Action.
  • Agent executions triggered by your own code via the Agent API. Those are billed as API calls against your agent's own quotas, not as AI Actions.

Test Runs​

A Test Run is one evaluation of one test case × one model.

  • Running 10 test cases against 3 models = 30 Test Runs
  • Running the same test case twice = 2 Test Runs
  • Re-running only failed cases counts each re-run individually

Test Runs require the Testing Framework, which is available on Pro and above.

Projects​

Caps on how many active projects an organization can hold. Archiving a project does not currently free a slot — delete old projects if you need room.

Members​

For Team and Enterprise plans, the member cap is typically expressed as seat count (and billed per seat). Invitations to pending members count toward the cap.

Storage​

Storage limits cover prompt content, test cases, test run history, and agent logs. Old test run data can be pruned from the workspace settings.

What happens when you hit a limit​

  • AI Actions exhausted: new AI Actions are blocked until your monthly reset (1st of the month) or until you upgrade.
  • Test Runs exhausted: the same — blocked until reset or upgrade.
  • Free tier daily cap hit: AI Actions pause for the rest of the day and resume automatically after 24 hours. The monthly allowance is not affected.
  • Projects / members / storage cap hit: you'll be prompted to upgrade or free up space before creating new resources.

For the exact numeric quotas on each tier, see the live pricing page.