Credits and usage
Quanty measures AI work in credits. Your plan gives the organization a monthly allowance; one-time top-ups add more. Everything that uses AI, from a chat answer to reading a document, draws from the same balance, so there is one number to watch. Admins can cap the spend of the whole organization and of each member.
Before you start
- Everyone sees their own credits on Settings, Usage. Admins see the whole organization in money too, the Billing page and Budget & limits.
- AI work needs a paid plan with credits (see Plans and billing).
What a credit is
One credit is $0.001 of AI work at Quanty's rates. A rate is what the provider charges for the work, plus 30 %: the same multiple for every model, for document search and for speech-to-text in voice input and AI columns. Two kinds of work have their own rates, shown in the table below: live web research, and transcribing call recordings beyond the hours your plan includes. A credit is not a model "token": the rates below are per million model tokens, and a credit is a unit of money.
All models run on Amazon Bedrock. Rates per million tokens, input (what the model reads) and output (what it writes):
| Model | Input / 1M | Output / 1M |
|---|---|---|
| Claude Fable 5 and Fable 5.1 | $13.00 | $65.00 |
| Claude Opus 5, Opus 4.8, Opus 4.6 | $6.50 | $32.50 |
| Claude Opus 5.5 | $5.20 | $26.00 |
| Claude Sonnet 4.6 | $3.90 | $19.50 |
| Claude Sonnet 5, Sonnet 5.5 | $2.60 | $13.00 |
| Claude Haiku 4.5 | $1.30 | $6.50 |
| Claude Haiku 5.5 | $0.13 | $0.65 |
| Amazon Nova 2 Lite | $0.507 | $4.251 |
| Amazon Nova Lite | $0.101 | $0.406 |
| Amazon Nova Micro | $0.060 | $0.239 |
| gpt-oss 120B (open weights) | $0.26 | $1.03 |
Claude Haiku 5.5 charges a request whose prompt is longer than 100 000 tokens at five times these rates, for the whole request.
Some models run through AWS's EU profile, which keeps the request inside EU Regions; by default these are Claude Opus 5, Opus 5.5, Sonnet 5 and Sonnet 5.5. AWS charges 10 % more for that, so their rates are 10 % higher than in the table (Sonnet 5: $2.86 and $14.30). Each column's settings show what it costs per row with each model (Cost per model, see Columns and AI).
What spends credits
| What | How it is charged |
|---|---|
| AI columns, chat, agents, dashboards and presentations made by AI, AI formula drafting | The model's tokens at the rates above. |
| Reading documents (scans and images) | The document analysis model's tokens; text files and spreadsheets are read at no cost. |
| Live web research | About 13 credits per search and about 3 per page read; a deep research run from about 125 credits, at most 5 000 per run. |
| Quanty Connect sources | About 13 to 75 credits per call, on top of the research run. |
| Transcribing audio and video (voice input, AI columns) | 13 credits per minute of audio. |
| Transcribing call recordings | Free up to your plan's included hours; after that about 900 credits per recorded hour. The AI summary of each recording is charged on top, at the model's rates. Record this call? shows the current figure. |
| AI Secretary calls | Once per call: speech recognition, the model, the voice and the phone gateway, plus 30 %. |
| WhatsApp messages and SMS | Per message, plus 30 %: to a Polish number about 7 credits (WhatsApp free text), 29 (WhatsApp template) and 42 (one SMS part). |
| File storage | 60 credits per GB per month. |
Free: formulas, lookups, the KSeF invoice data column, cells a branch skips, imports and exports, and sign-in codes by SMS. See also What consumes credits? on the Billing page.
Before a large run, an AI column shows the estimate ("About 240 credits for 60 cells."), and the run confirmation shows your balance (see Columns and AI).
Your balance
The AI credits card on Billing shows the credits left, "120 000 / 315 000 used this period", the Top-up balance and when the allowance renews. The balance has two parts, spent in this order:
- The monthly allowance of your plan, granted every month. What is left at the end of the month does not carry over.
- Top-up credits, bought as one-time packs. They carry over from month to month and are used only after the allowance.
When the subscription ends, the rest of the allowance expires and the top-up credits stay.
Warnings appear by themselves:
- "Running low: less than 10% of your monthly allowance left"
- "Credits used up. AI features are paused": AI stops until you top up or the month renews. Everything else keeps working.
Top-ups
Top up credits on Billing: "One-time packs. Top-up credits roll over and are used after your monthly allowance."
| Pack | Price |
|---|---|
| 6 750 credits | $9 (32 zł) |
| 18 000 credits (Most popular) | $25 (90 zł) |
| 54 000 credits | $75 (270 zł) |
Pick a pack and click Buy; Stripe's checkout opens. When the payment is through, the credits are in your balance at once ("Payment received"), with a notification "Top-up credited" and a Stripe invoice. "An active subscription is required to buy top-ups."
Storage
File storage on Billing shows the space used of your plan's (5, 25 or 100 GB): "Everything in Files, chat attachments and sheet cells counts towards the plan storage." Storage packs add space once and for good: +10 GB for $5, +50 GB for $19, +250 GB for $69 (Buy storage).
Recent credit activity
Recent credit activity on Billing lists every movement of the balance with Date, Type, Credits and Category:
| Type | What it is |
|---|---|
| Plan allowance granted | The plan's credits for the month. |
| Period reset | The new month's allowance. |
| AI usage | Credits spent by AI work. |
| Top-up purchase | A pack you bought. |
| Storage pack purchase | A storage pack you bought. |
| Adjustment | A correction by Quanty. |
| Allowance expired | The rest of the allowance after the subscription ended. |
Budgets and spend limits
Settings, Organizations, Budget & limits (admins):
| Setting | What it means |
|---|---|
| Seats | "Occupied seats (active members + pending invitations) out of the plan limit." |
| Monthly AI budget | "Caps AI spend for the whole organization. This organization is billed for its own usage." Spent this month and the cap, in US dollars, or No limit. Edit changes it. |
| Member AI spend limits | "Monthly cap on each member's AI usage. Leave empty for no limit." Each member's AI spend / limit, with Set limit. |
Set limit opens Monthly AI spend limit: "Piotr Spedytor has spent $0.00 this month. Set a monthly cap in USD, or leave empty for no limit." Admins can cap anyone, other admins included.
When a cap is reached
- Budgets cover the calendar month: "Usage counters and budgets cover the current calendar month and reset on the 1st of each month (UTC)."
- At the organization's cap, AI work stops for everyone until the 1st or until an admin raises it: "has reached its monthly AI budget. An organization admin can raise the budget in organization settings."
- At a member's cap, AI stops for that member: "You have reached your monthly AI budget". "An organization admin can raise your limit."
- The member and all admins get a notification ("Organization AI budget reached", "Monthly AI usage limit reached"), once a month, by e-mail too if their alerts say so. There is no warning before the cap.
An automation that runs AI columns pauses when the budget runs out (see Testing and run history).
The Usage page
Settings, Usage:
Admins see:
| Part | What it shows |
|---|---|
| Header | The date range, Export CSV (every AI operation with its time, person, category, model, tokens and cost) and Manage plan. |
| Your spend, Organization spend | Money spent in the period. |
| Limit alerts | "How to tell you when your monthly AI budget or the organization budget is reached": In-app notification, Email alert and the Alert email address, with Save alerts. |
| Budgets | Your monthly limit, Organization monthly budget, Spent this month (org), when they reset, and Manage budget & limits. |
| AI credits, Team members, File storage | Used of the plan. |
| Chat & calls | Call hours and Included transcription this month (see below). |
| Chat, Library, Web search, Other AI | Spend by area; Other AI splits into categories such as Transcription (voice input and calls), Presentations, Dashboards, Automations, WhatsApp messages and SMS. |
| Members and models | Each member's usage, cost and Monthly limit, and each model's tokens, cost and operations. |
Members and viewers see "Your AI credits this period" ("Credits you have used yourself since the period started.") and the Organization credit pool: "Credits are shared across the whole organization. Prices and invoices are handled by its administrators."
Call hours
Team calls in Chat use the plan's call hours: per user per month (20 h on Silver, 80 h on Gold, 160 h on Platinum), multiplied by the plan's seats and shared by the organization: "20 h per user × 2 seats". Chat & calls on Usage shows what is left and when it resets. "Time everyone spends in calls counts."
- Admins are notified at 80 % ("Call hours running low") and when they are used up ("Call hours used up").
- When the hours are used up, a call in progress continues, and new calls are refused until the month renews.
- Included transcription works the same way for transcribing call recordings: "After that, AI credits are used."
- Call hours are never unlimited, on any plan. See Calls and recordings.
Good to know
- One pool. Credits belong to the organization, not to a person; a member's limit only caps their share.
- Estimates. The costs above are approximate: a longer document, prompt or answer costs more. The real cost of every run is on Usage and in each cell's details (see Cell details, reviews and history).
- Price cuts are passed on, and changes to how fast credits are spent are announced at least 30 days ahead (see Plans and billing).
Related pages
- Plans and billing: plans, prices and the subscription.
- Columns and AI: what an AI column costs before you run it.
- Organizations and roles: who is an admin.
- Testing and run history: what automations cost.