Docs › Writing Studio › AI cost & quota
AI cost & quota
Last updated May 15, 2026 · 4 min read
AI in Slima isn't unlimited — each operation consumes credits. This article tells you how much was used, how much is left, and what to do when you're at zero. No exact numbers — pricing / formulas live at /plans, which evolves.

See your balance
Three places:
- The pre-action estimate dialog — before a credit-costing op: "~X credits / you have Y left"
- Account → Usage — the fullest view (below)
- Account → Overview — a card with this week's AI usage summary
The sync indicator next to the book title does not show credits — credits live only in the places above.
Account → Usage: three tabs
- Purchase Credits — top up
- Usage Stats — AI usage, permanent credit, report quota
- History — every credit earned / spent, with timestamps
Estimates before you spend
Before an action that costs credits, Slima usually:
- Estimates the cost
- Shows "Estimated X credits / you have Y left"
- Waits for your confirm
E.g.:
- "Rewrite this chapter" → a confirmation dialog: "~X credits (you have Y)"
- Large ops (chapter analysis, chapter rewrite) almost always prompt to confirm
- Small ops (single-sentence rewrite, quick action) just run
Out of credits: two quota prompts
Two surfaces:
The pre-action confirmation dialog
"This operation estimates X credits; you have Y"
Depending on your account state:
- Start Free Trial (if you haven't used it) or View Plans — go to /plans
- Cancel the operation
Running out mid-operation
Rare — operation exceeded estimate mid-flight:
- What's been produced stays
- the confirmation dialog opens for the three options above
Why we don't list numbers here
This article deliberately doesn't say "X credits per chat" or "Pro gives Y per month" because:
- Models / pricing adjust
- Tokens per chat vary wildly with context
- Hard-coded numbers in docs mislead within months
Numbers live at slima.ai/plans — always current. Inside Slima, Account → Usage shows your live state.
Credit-saving habits
- Use quick actions for small edits (Quick actions on selection) — cheaper than a full chat
- Web search off when not needed (Web search) — extra cost when on
- Split long chats — 50+ message chats compress but still rack tokens; new topic → new chat
- One decisive big edit instead of four nudges — don't quick-action the same passage 4 times; ask for "rewrite, give me 3 versions" once