1 hour ago
16
Forkast did the math on Visa/Artemis: an average x402 payment of $0.14 across 109.6 million adjusted transactions. That’s adoption, but of micropayments — not baskets, bills or remittances.
The $0.14 average that demotes “adoption” to micropayments
Visa, using Artemis Analytics, says the agent rail has logged roughly 109.6 million adjusted transactions settling about $15.0 million in value as of April 21, 2026. This is not testnet noise. It’s a production rail with real users and real on-chain receipts, just at tiny ticket sizes.
Separate work from Chainalysis shows the scale of counts. x402 activity on Base went from near-zero in mid-2025 to more than 100 million cumulative transactions by the end of Q1 2026. A late-2025 burst came from a pay-to-mint meme experiment called PING, which pushed roughly 150,000 transactions in its first month and briefly sent weekly x402 traffic up by more than 10,000%.
When raw counts mislead: filters, experiments, and manufactured flow
Headline counts flatter the story. The most useful numbers are the adjusted ones. A forensic read of Visa–Artemis by The Legal Futurist reconciles the raw and the filtered figures: about $135.7 million across ~178.3 million raw x402 transactions shrank to $15.0 million across 109.6 million after Artemis applied heuristics. In other words, roughly 89% of reported dollar volume and about 39% of transactions looked like wash or testing rather than organic spend.
Measure Raw (approx.) Adjusted Removed (%) Transactions ~178.3M 109.6M ~39% Dollar volume ~$135.7M ~$15.0M ~89%
Chainalysis’s PING episode shows how those raw spikes get manufactured. Pay-to-mint incentives create a flood of transactions that prove throughput capacity but don’t establish steady willingness to pay at meaningful size. The gas and micro-fees are real; the economic signal is faint.
Three measurement pitfalls sit behind the “big number” headlines:
- Wash/test loops that never intend to transfer durable value
- Pay-to-mint farming that rewards clicking, not spending
- Registration churn that appears as activity without commerce
Once you strip those out, you’re left with production traffic centered on very small payments. That inflates blockspace usage and dashboard metrics but doesn’t imply rising token demand.
Heavy usage, light tokens: PING as a value‑capture failure
The project that helped ignite the surge didn’t capture it in price. CoinGecko shows PING at roughly $0.0011 with a market cap near $1.1 million (accessed Aug 20, 2026). That’s a rounding error next to its on-chain footprint. Pay-to-mint mechanics reward transacting. They don’t guarantee cash flows or claims that accrue back to the token.
Agents registered, economies missing
Even where agent infrastructure is shipping, the economic plumbing is thin. An arXiv study of the first 10,000 ERC‑8004 agents on Ethereum finds heavy identity and registration, but limited metadata, service exposure, reputation, and transfer activity. In plain terms: many agents exist on-chain in name; few show signs of rich commercial interactions.
So the rail can mint agent accounts at scale. What it hasn’t shown yet is sustained payment size, recurring service delivery, or the kind of fee flows that justify repricing native or related tokens.
Same surge, different economy: cross‑chain contrasts
Agent and machine traffic is not limited to one stack, but the money concentrates differently by chain. Nansen’s NEAR Q4 2025 report tallies about 54 million transactions and roughly 20 million users in the quarter (around 4.2 million daily transactions on average), with 8.4% quarter‑over‑quarter transaction growth. More telling is money type: Tether usage on NEAR rose 29% QoQ. That hints at where real spend may pool first — stablecoin rails tied to concrete payment use — versus meme‑oriented agent flows that punch above their economic weight.
Chains capturing stablecoin throughput can still show high counts, but the cash leg is clearer and bigger. Agent rails that skew toward farmed or experimental flows mostly add clicks.
And the constraint is hard. With about $15 million settled across 109.6 million adjusted agent transactions — and most raw dollar volume stripped as non‑organic — there isn’t enough value per click for tokens to rerate unless payment sizes lift or persistent stablecoin spend shows up on these rails.
Disclaimer: This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial, or other advice.
English (US) ·