State of Agent Commerce — August 2026
In August 2026, AI agents settled about $49.9k USDC a day on x402 (Agent Settlement Volume, a 7-day rolling average of daily medians, covering all settlement Paddock observes). About 9,456 transactions a day were tied to a named service (Attributed Volume, ~84.8% of the total). State of Agent Commerce, Paddock, published 3 Sep 2026.
9,456 attributed txns/day [Attributed Volume, ~84.8%] (of the 841 recipient wallets captured on 2026-08-31, top-N by transaction count, list_complete: false) is the stable read for August. That figure is attribution-tested and scope-disclosed, use it when you need a defensible number. ASV stands at $49.9k USDC and 1,222,064 txns/day [Total Settled Volume], 7-day rolling average of daily medians, total settlement Paddock observes via x402scan's all-facilitator aggregate (mixed/low-confidence basis, see methodology note), which sets the ceiling: it captures everything Paddock observes, including unattributed and circular-flagged volume. The gap between the two figures is not noise, it is the honest distance between "all transactions observed" and "transactions tied to a known service."
Showing top 5 of all categories. Full breakdown in the paid report. Of the 841 recipient wallets captured on 2026-08-31, top-N by transaction count, list_complete: false.
Transaction volume grew 680% between 2026-08-01 and 2026-08-31 (two-point comparison, not a 29-day average).
You're reading the free preview of the August 2026 issue. The full report — every section, the complete analysis and data — is part of the monthly subscription.
Subscribe to read past issues →Query the entire agent economy in plain English.
Every number in this report is live in the Paddock API — and you don't need a query language to reach it. Connect Paddock's MCP server to Claude, ChatGPT, or any LLM client and ask:
Paddock runs the question against the full snapshot archive and attribution graph and hands your model back clean, cited data — so your next memo, model, or dashboard runs on numbers no one else has.