Abstract
This HIP is the canonical chronology of the Hanzo network: the four locked version milestones, their activation dates, the AI compute infrastructure at each step, and what the triumvirate (DEX + EVM + FHE) came to mean for the network. Hanzo AI Inc. (Techstars '17) shipped 1.0 as a centralised AI compute network and graduated through PQ and GPU-native to a sovereign Lux-family L1 at 4.0. This document is the single source of truth for that arc.
Motivation
Hanzo's version history spans five years and four quite different infrastructure designs. Without a canonical chronology, partners and integrators are forced to reconstruct the timeline from papers, HIPs, and changelogs. This HIP fixes the timeline.
Chronology Table
| Version | Active from | Theme | Defining changes | Reference paper |
|---|---|---|---|---|
| 1.0 | 2018--2023 | Initial AI compute network | Centralised inference and training infrastructure with off-chain billing. | hanzo-network-whitepaper (historical) |
| 2.0 | 2024 | Hamiltonian LLMs + AI Coin | On-chain accounting for inference; AI Coin native asset; first version of HMM (Hanzo Market Maker, HIP-8); ASO routing protocol (HIP-2). | hanzo-ai-chain, hanzo-tokenomics, lightspeed-dex |
| 3.0 | 2025 | Full PQ | ML-DSA mandatory for validators; Hanzo settles under Lux Quasar 3.0 (LP-105) cert lanes; FHE inference research preview. | hanzo-pq-crypto, hanzo-fhe-inference |
| 4.0 | 2026-02-14 | GPU-native + native DEX with | Sovereign L1 on the Lux-family primary chain template (LP-134); AIVM GPU-resident with 8 precompiles (0x0a01--0x0a08); HMM cross-listed via ; FHE inference promoted to production; Proof-of-AI mining. | hanzo-4-0-launch |
4.0 Triumvirate
A Lux-family L1 is triumvirate-complete when it carries three first-class subsystems. Hanzo 4.0 specialises each one for AI compute:
- DEX --- HMM (Hanzo Market Maker). The native AI-compute
exchange. Five compute primitives traded: inference seconds,
training slots, GPU leases, embedding throughput, and active-inference
budgets. Capacity-decay AMM curve. Cross-listed with Lux D-Chain
and Zoo DEX via (LP-310). Registered at
h_0 + 12. - EVM --- AIVM. A Cancun-compatible EVM with eight AI-specific
precompiles in the
0x0a01--0x0a08range:HMM_LOOKUP,INFER_REQUEST,INFER_VERIFY,TRAIN_BOUNTY,GPU_LEASE,AGENT_AUTH,F_CALL,M_CUSTODY. All GPU-resident at the 4.0 boundary. - FHE --- Confidential AI inference. Encrypted prompt inference,
encrypted decision-tree inference, encrypted federated aggregation.
Reached through the AIVM
F_CALLprecompile; threshold decryption delegated to Lux F-Chain when the client prefers Lux-rooted custody.
Tokenomics Continuity
The AI Coin (AI) supply schedule from 2.0 is preserved at 4.0. The
4.0 boundary added two emission streams entirely from the
previously-allocated ecosystem-incentives bucket:
- HMM fee routing: 0.30% of swap volume (50/30/15/5 split).
- FHE relay rewards: capped at 0.5% annualised; paid to validators performing FHE bootstrapping and ciphertext maintenance.
No new emission was added.
Activation Receipts (4.0)
- Activation block:
h_0 = 36,963,000. - Activation timestamp: 2026-02-14T00:00:00Z.
- Active validators at
h_0 + 24h: 312. - Aggregate GPU FLOPs (FP16): 4.2 EF.
- Sustained inference TPS: 2.1M.
- Median LLM-7B inference latency: 42 ms.
- Median HMM swap latency: 0.91 ms.
References
- HIP-2 ASO (Active Semantic Optimization) protocol
- HIP-7 Active Inference Integration for Hamiltonian LLMs
- HIP-8 HMM (Hanzo Market Maker) --- Native DEX for AI Compute Resources
- LP-009 GPU-Native EVM
- LP-010 v4 QuasarSTM 4.0
- LP-013 v2 F-Chain FHE Service
- LP-105 Quasar 3.0 Consensus
- LP-134 Canonical 9-Chain Topology and Primary Chain Templates
- LP-137 GPU-Residency Invariant for Lux-Family L1s
- LP-310 Common Settlement API
hanzo-4-0-launch(papers repo)hanzo-ai-chain(papers repo)hanzo-fhe-inference(papers repo)lightspeed-dex(papers repo)hanzo-tokenomics(papers repo)
Status
Final --- as of 2026-02-14, this chronology is locked. Any future major version (5.0, etc.) will require its own chronology proposal.