What are credits and how do they work?
Credits are the single unit that powers everything automation and AI inside Superchat.
Each time an automation runs or an AI Agent performs an action, credits are consumed. This includes:
Automation nodes – each executed node (e.g. Send Message, Assign Conversation, Add Internal Note) uses 1 credit
AI Agent responses – each AI-generated message uses on average around 7 credits, depending on the context, knowledge size, and message complexity.
AI Agent tool calls – actions like writing notes or updating contact data use a few additional credits
Important: Credits are only consumed for automation runs that actually execute and "do something." If an automation is triggered but stops before reaching an action node (for example, a trigger → filter → send message flow where the filter stops the run), no credits are used. If the automation continues to the Send Message node, then 2 credits are used (one for the filter and one for the send action — triggers do not consume any credits). This way, you only pay when something is actually automated.
How many credits do I get with my plan?
Plan | Credits |
Free | 500 |
Basic | 1,500 |
Professional | 7,000 |
Advanced | 25,000 |
Individual | Custom |
How can I manage my credit balance?
Every Superchat plan includes a monthly credit balance. If you need more credits, you can purchase a Credit Add-on to increase your monthly balance. You can do this easily under Settings > Plan & Billing > Credits — just click "Upgrade" to buy additional credits.
Both add-on and included credits renew automatically on the first of each month. Unused credits expire at the end of each month.
An increase in credits is only possible as an add-on — a one-time increase is not available.
When and how will I be notified about credit limits?
To help you manage usage, Superchat automatically emails Admins when credit usage reaches 80%, 90%, and 100%.
When all credits are used up, all automations pause automatically. Automations paused due to a lack of credits are clearly marked as "Paused – No Credits" so you can easily distinguish them from manually paused ones.
There is no queuing system, meaning automation runs that were stopped because credits ran out cannot be resumed — they are permanently stopped. Once you have credits again, new automation runs will start normally.
You can either upgrade your plan, add a monthly credit add-on to your subscription, or wait until the next month's reset to resume automations. Always make sure you have enough credits available to avoid interruptions.
How can I track and understand my credit usage?
You can monitor all activity in Settings → Plan & Billing → Credits, where you'll see:
Total credits used in a selected time period
Split between Automation Tasks and AI Agent
A breakdown by Automation and AI Agent, plus real-time usage trends.
For each automation run, you can also view:
Total credits consumed per run
Total credits consumed per node
This gives you full transparency and control across your workspace.
What affects my AI Agent's credit usage?
The number of credits an AI Agent consumes depends mainly on:
Prompt size and complexity — longer or more detailed prompts use more processing.
Knowledge base size — the more information the AI Agent uses, the higher the cost.
Conversation context and richness — long message histories, multiple attachments, or complex data increase processing.
In short: shorter prompts, focused conversations, and well-structured knowledge help your AI Agent work more efficiently.
How can I reduce my automation credit usage?
Since each executed action node consumes a credit, there are a few simple ways to reduce unnecessary usage:
Place filters early in your automation flows — if a run is stopped by a filter before reaching an action node, no credits are consumed. Moving filters earlier means unnecessary runs are stopped before they cost anything.
Regularly check and deactivate unused automations — automations that are still active but no longer needed will unnecessarily consume credits. Regularly review your automations and deactivate any that are no longer in use.
Review high-frequency automations — automations triggered by high-volume events (such as every new incoming message) can consume a large number of credits quickly. Consider if the triggers you have selected are the most suitable for your desired outcome.
Test automation variability yourself — before setting up your automation for public use, consider testing as many variations as possible, to ensure that the automation behaves as expected and there aren't any issues that will use up unnecessary credits later on. To learn more about testing your automation, click here.
How can I reduce my AI Agent's credit usage?
Keep prompts short and literal — vague instructions like "be helpful" or "respond professionally" don't give the AI Agent anything concrete to work with and add unnecessary processing without improving its responses. Instead, write specific instructions that describe exactly how you want it to behave, for example "always respond in English" or "do not discuss competitor products".
Separate behavior and knowledge — Prompts define how your AI Agent acts; the knowledge section defines what it knows; conflating the two makes your AI Agent less efficient.
Use Superchat QA lists (a simple list of relevant questions and their corresponding answers) — They're the most reliable and cost-efficient data sources for your AI Agent to process.
Use structured knowledge — Avoid uploading documents with complex formatting, tables, or images — the AI Agent can only process plain text, so cluttered documents add unnecessary processing.
FAQ
Does the AI Agent Playground consume credits?
No. Testing your AI Agent in the Playground is completely free and does not use any credits.
Is the AI Copilot feature part of the credit system?
Not yet. The AI Copilot is currently seat-priced, but it will also be included in the credit system in the near future.
Can I set a credit limit per automation?
No. It is not currently possible to set a credit limit on individual automations. The best way to control usage is to review your automations regularly, deactivate any that are no longer needed, and to make sure filters are placed early in your flows to prevent unnecessary runs.