Cost by plan as AO volume grows
Plan rates held at your current inputs β only monthly AO volume changes per row. Free plan shows "β" once its 1,000 AO hard cap is exceeded (no overage available).
| AOs/mo | Free | Individual | Team | Cheapest valid |
|---|
Agent Operations: a cost-per-outcome meter, not a data-volume meter
Classic Airbyte Cloud bills by data moved: credits at $2.50 each, burned by GB synced (database sources) or rows synced (API sources) β see the ETL / Data Pipeline Cost Calculator for that model. Agent Operations is a separate, newer product line built for AI agents that continuously loop through Search, Read, Act and Reason steps rather than running a single batch sync β so Airbyte meters each loop step as one Agent Operation (AO) instead of counting bytes or rows. That's a genuine "cost per outcome" framework: an agent that does more work per task burns more AOs regardless of how little raw data it actually moves.
The self-serve tiers are flat monthly plans with an included AO quota plus overage: Free ($0/mo, 1,000 AOs, daily Context Store refresh, no overage β the workload simply stops past quota), Individual ($29/mo, 5,000 AOs included, hourly refresh, $0.004/AO overage) and Team ($299/mo, 10,000 AOs included, multi-user workspaces with parallel workflow execution, $0.005/AO overage). A Custom/Enterprise tier above Team adds a dedicated Solutions Architect and SLA-backed support at negotiated pricing. Because Individual's overage rate is actually cheaper per unit than Team's, teams typically upgrade to Team for the multi-user and parallel-execution features rather than because the raw AO math demands it β model both below at your expected volume.
How this calculator works
The Airbyte Agent Operations Cost Calculator models Airbyte's AI-agent billing line across its three self-serve plans at your expected monthly AO volume: Free (flat $0, capped hard at 1,000 AOs with no overage β shows "β" past that), Individual ($29/mo base + $0.004 per AO beyond the included quota) and Team ($299/mo base + $0.005 per AO beyond its included quota). All three totals are shown side by side along with which plans are actually usable at your chosen user/workspace count (Free and Individual are single-user only), then a second table isolates every plan across a range of AO volumes at your current rates so you can see exactly where Individual's overage crosses into Team territory.
These are representative figures compiled from Airbyte's public 2026 Agent Operations pricing page and third-party pricing breakdowns, not a live vendor feed β Airbyte can adjust AO quotas, overage rates and features, and Custom/Enterprise pricing above Team is negotiated directly with sales. Every field is editable: enter your own expected AO volume, plan quotas and rates for a number that matches your actual workload.
Frequently asked questions
What is an Airbyte Agent Operation (AO)?
An Agent Operation (AO) is Airbyte's metering unit for its Agent Operations product line β a separate billing model from classic Airbyte Cloud's row/GB credits. Airbyte describes AI agents as looping continuously through four cognitive steps (Search, Read, Act, Reason) rather than doing a single batch sync, so it meters each of those loop steps as one AO instead of counting rows or gigabytes moved. That makes AOs a "cost per outcome" unit β you pay for how much an agent actually did, not how much data passed through a connector.
How is Agent Operations pricing different from Airbyte Cloud's credit model?
Airbyte Cloud (classic ETL) sells credits at a flat $2.50 each and burns them by data volume β roughly $10/GB for database sources or $15/million rows for API sources. Agent Operations instead sells flat monthly plans with an included AO quota and per-AO overage rates: Free (1,000 AOs, no overage), Individual at $29/mo (5,000 AOs, $0.004/AO overage) and Team at $299/mo (10,000 AOs, $0.005/AO overage). The two product lines are billed on completely different meters β don't assume your existing Airbyte Cloud connector spend predicts your Agent Operations bill.
How many AOs does a typical AI agent workload use per month?
It depends heavily on how many context lookups and actions the agent chain performs per task, but publicly reported estimates put a typical workload syncing five data sources on an hourly cadence in the $200-500/month range once past the free tier β meaning several thousand to tens of thousands of AOs a month for a modest single-team deployment. Agents that re-query context on every turn of a long conversation burn AOs much faster than ones that cache results, so the biggest lever is how aggressively you cache Context Store lookups rather than raw request volume.
When does the Team plan beat paying overage on Individual?
Purely on per-AO cost, Individual's $0.004 overage rate is cheaper than Team's $0.005 β up to the point where the fixed cost difference ($270/mo) outweighs it, around 54,000 extra AOs beyond each plan's included quota, the two plans cost the same. But Team is also the only tier with multiple users, workspaces and parallel workflow execution, so most teams upgrade for those features long before the raw AO math forces it, not because Individual becomes more expensive.
Is Airbyte's Free Agent Operations plan usable for production?
Only for very light usage or evaluation β Free includes 1,000 AOs/month with daily (not hourly) Context Store refreshes, a single user, and no overage option at all, so once you exceed 1,000 AOs in a month the workload simply stops rather than billing you extra. Any agent running on a real schedule against live data will typically need Individual at minimum for hourly refresh and overage headroom.
Are these exact current Airbyte Agent Operations prices?
These figures are compiled from Airbyte's public 2026 Agent Operations pricing page and third-party pricing breakdowns, not a live vendor feed β Airbyte can and does adjust AO quotas, overage rates and plan features, and Enterprise/Custom pricing is negotiated directly with sales. Every field below is editable: plug in your own expected AO volume and the calculator recalculates all three self-serve tiers instantly.