Credits & Billing

Atanzo AI uses a credit-based model to track usage across guided sessions, agent automation, and embedded SDK deployments. This article explains how to view your usage and what activities consume credits.

Viewing Your Usage

Navigate to Settings › Credits at /app/settings/credits. This page shows:

  • Your current credit balance
  • A usage history log
  • Credit consumption broken down by session or activity

What Consumes Credits

The following activities draw from your credit pool:

  • Guided task sessions — each voice or video session between a worker and an agent.
  • Agent runs — automation triggered during a guided task or independently as a scheduled or event-driven workflow.
  • Camera input and visual identification — image capture and processing during a session (for example, identifying a part or reading a label).
  • SDK-embedded sessions (only if you have integrated Atanzo AI into your own application) — these draw from the same credit pool as sessions started in the console.

What Drives Your Credit Usage

The cost of a session isn't a single flat number — it's driven by two things happening at the same time:

  • Time in session — charges accrue for each minute a session stays open, covering the infrastructure that keeps the worker connected to the agent.
  • Exchanges with the agent — each back-and-forth with the AI, including any camera-based identification during the session, adds to the cost on top of the time-based charge.

A small platform margin is built into the credit cost of every session. You don't need to calculate or track this separately — it's already reflected in the totals shown in your usage history.

Because cost scales with both session length and how much the worker interacts with the agent, a short session with a lot of back-and-forth can end up costing more than a longer, mostly-listening one, and vice versa.

If Credits Run Out Mid-Session

If your credit balance runs out while a session is active, Atanzo AI ends the session gracefully — the worker is brought to a clean stopping point rather than cut off mid-step. The session record is preserved, so whatever guidance was delivered up to that point stays available in the usage history.

Credit Caps Per Session

Per-session credit caps are an SDK capability, available to teams embedding Atanzo AI in their own application. Your backend can pass a maxCreditsPerSession value when minting a session credential — see Getting a Session Credential for details. This is useful for preventing runaway usage in high-volume or automated deployments.

For sessions started directly in the Atanzo AI console, there isn't a per-session credit cap you configure yourself in the admin UI today. If you need usage limits on console-managed sessions, contact your account manager.

How to Keep Usage Predictable

A few habits make credit usage easier to forecast:

  • Keep sessions focused. Shorter, well-scoped sessions cost less than long ones, since time-based charges accrue for the whole time a session stays open — including pauses.
  • Use the camera only where it earns its place. Turn on camera input for tasks that genuinely need visual identification; voice-only sessions cost less than sessions running camera input throughout.
  • Cap embedded deployments. If you're integrating Atanzo AI into your own application, set a maxCreditsPerSession value (see Credit Caps Per Session) so no single embedded session can run away with usage.

Managing Your Plan

To change your plan or purchase additional credits, contact your account manager or visit the billing section in your account settings.