AI features in ComplyWise draw from a dedicated AI credit pool, separate from your screening credits. This reference lists what spends credits, what is free, and how to read your balance.
What is metered
| Action | Cost |
|---|---|
| Chat reply | 1 credit x the model's multiplier |
| Improve prompt | 1 credit |
| Per-document AI review | 1 credit x the model's multiplier per run |
| Review-table fill | 1 credit x the model's multiplier per document call |
| DDQ answer drafting | 1 credit x the model's multiplier per batched call |
| DDQ question extraction | metered per import |
The model multiplier is shown in the model picker: a x1 model spends one credit per action, a x3 model spends three.
What is not metered
- Clause search: unmetered, always.
- Failed calls: an error, an unreadable scan, or a stopped provider response charges nothing; the UI says No credits used.
- Reading, flagging, and commenting on reviews; editing and approving DDQ answers; managing prompts, skills, and projects; downloading or saving artifacts.
- KYB Manager document extraction and verification: those run outside the AI credit pool.
- Interrupted work: a chat reply that never finished, marked Interrupted in the thread, and a document review that timed out both charge nothing.
Credits are consumed only when a call succeeds. Stopping a streaming chat reply keeps the partial answer and still counts as that reply's credit.
Reading your balance
- The credit chip near the Copilot composer shows credits used versus your limit, with the cost of heavier models surfaced in tooltips.
- At 80% of the pool, workspace admins get a warning notification; at 100% another. A monthly pool warns again in each new period; a total pool warns once.
- At zero, AI actions are disabled with a limit-reached notice. Nothing breaks: documents, reviews, and DDQs remain readable.
Monthly vs total pools
Your pool is configured one of two ways:
- Monthly: the limit resets each month.
- Total: one fixed allowance with no reset.
Some workspaces run unmetered, with no limit set. Your platform team sets the pool and its size; e-signature envelopes and screening runs have their own separate pools and never touch AI credits.
Adjusting your limit
Credit limits are set per workspace by the platform team. To raise a limit, change the pool type, or question a charge, contact support@complywise.io.
Related articles
- What the Legal AI Copilot does: capabilities and privacy
- Chat basics: the model picker and multipliers
- Review tables: how per-document calls are counted