Autonomous AI agents now drive 38% of on-chain crypto volume. This indicator is the new M2: it measures the liquidity actually deployed by autonomous entities, without human bias.
Last AI verdict update : April 28, 2026
MacroView's AI detects a parabolic agent-to-agent adoption phase. The sector's market cap multiplied 11x in 12 months, but the volume/cap ratio stays at 0.42 — well below the bubble zone (1.2+). The sector is in healthy expansion, not mania. Three sub-categories outperform: ASI Alliance (FET/AGIX/OCEAN merged), high-throughput L1s (SOL, SUI), and AI middleware (LINK, GRT). Recommended institutional allocation: 8–12% of crypto book.
% of Ethereum + Solana volume — 2024 → 2026
In $B — Top 25 agent infrastructure tokens
| Token | Category | Market Cap | YTD 2026 | AI Score |
|---|---|---|---|---|
| ASI (FET fusion) | Agent Infrastructure | $48.2B | +312% | 92/100 |
| SOL | High-Throughput L1 | $98.4B | +184% | 88/100 |
| LINK | Oracle Middleware | $32.1B | +156% | 85/100 |
| SUI | Move-based L1 | $18.7B | +228% | 83/100 |
| RNDR | GPU Compute | $12.4B | +267% | 80/100 |
| GRT | Indexing Layer | $8.9B | +142% | 78/100 |
M2 measures purchasing capacity in the traditional economy. But in crypto, purchasing capacity is no longer enough to explain flows: you need to understand deployed functional utility. That's what agent-to-agent volume measures.
When an AI agent pays another AI agent for a service (data, compute, inference), that's real economic value created and settled on-chain. No human in the loop. No emotion. No FOMO. Pure functional adoption.
The short history (since 2024) shows a correlation of +0.91 between agent volume growth and infrastructure token performance. This signal precedes rallies by 6 to 9 weeks — exactly like M2 precedes BTC.