Our team is actively looking into implementation options for Airtable’s AI - both in agents and in automation. However, we need to have better visibility into ongoing credit usage to estimate long-term costs and optimize for efficiency.
Although Airtable has an AI credit usage summary for Enterprise in the Admin Panel, that data is not robust or accessible enough to use for long-term planning. This is particularly a problem because our field agents and automations do not provide any data about the number of credits used or estimated usage. It’s also not accessible for most of the team, not allowing individual users to moderate their own usage to ensure we don’t hit our cap.
Ideally, AI credit usage would be tracked in the audit log, including (if available) the user that triggered the agent to run. Either way, our team needs a way to track, on a per-base, per-day basis the usage of AI credits so that we can test AI implementations in our bases without accidentally maxing out for the month. This data should be accessible via the API or in a way where we can share trend & usage data with non-admin users.
