Gyblos: an open market for intelligence

Litepaper, draft 0.4. A living document: it is updated as the protocol is built and every claim is marked as built, tested or planned. Not financial or legal advice.

Two kinds of "token" appear in this paper. Inference tokens are the units of text a model reads and writes. GYB is the provisional name of the Gyblos network token. When the paper says "intelligence", it means inference tokens.

Abstract

Access to AI is becoming access to opportunity, and it is unevenly distributed: frontier inference needs a credit card, an account with a lab and the right country. Meanwhile, idle capacity that could serve it sits in gaming PCs, Macs, servers and, soon, phones, TVs and cars.

Gyblos is a permissionless, peer-to-peer market where anyone can sell, buy or give away inference. Buyers pay per inference token in USDC on Solana through payment channels, keep control of their funds and get unused deposits back. Sellers set their own prices and pay a single 3% protocol fee. Trust comes from measurement, not identity: there is no KYC, every seller is an address with an automatic service reputation, and open-weight models can earn a cryptographic verified-model seal. A network token, GYB, rewards measured honest service from day one, locked and released gradually, and is never paid for raw volume.

1. Vision: intelligence as a commodity

Inference tokens are becoming a basic good, closer to electricity or bandwidth than to software. Whoever can produce them can sell them, and whoever needs them should be able to buy them in any quantity, from anywhere, without permission.

We expect production to spread to every device with a chip. Today the producers are GPUs in homes and data centers. Next are laptops and phones while they charge, and later TVs and parked cars. Each one can be a small producer of intelligence.

Gyblos is the market layer for that world, and it serves three kinds of participant:

• Those who sell. Any device owner can turn idle capacity into income.
• Those who buy. People and AI agents get intelligence with nothing more than a wallet and a few cents of USDC.
• Those who give. Producers can donate capacity to people, schools and communities that would otherwise have no access. People without access to AI risk becoming the illiterate of the future; a market that also lets intelligence be given away is part of the answer.

2. The problem

1. Access is gated. Cards, accounts, regions and minimum purchases exclude billions of people and every autonomous agent.
2. Idle supply has no market. Home hardware cannot easily sell: NAT blocks incoming connections, payments need a company and buyers have no reason to trust an anonymous node.
3. Buyers cannot see what they get. A Stanford audit (ICLR 2025) found 11 of 31 Llama endpoints serving outputs that differ from the reference weights. Routers and P2P markets mostly do not check.
4. Incentives get gamed. Networks that reward volume pay people to buy from themselves, and wash trading follows.

3. How Gyblos works

A session in four steps:

1. Discover. The proxy reads seller identities from the 8004 registry on Solana and fetches signed offers (model, price, region, data policy).
2. Open. The buyer locks a budget in a payment channel. The ceiling is enforced on-chain, so no seller and no prompt injection can raise it.
3. Use. Requests go straight to the seller; prompts never pass through Gyblos servers. Payment is metered per inference token, with signed vouchers (MPP sessions) or per call (x402 upto).
4. Settle. The seller settles on-chain; the program pays the seller, takes the fee and refunds the unused deposit in the same transaction.

Home sellers are first-class: connectivity through NAT uses hole punching with community relays as fallback (planned).

4. Trust without identity

Gyblos never asks who you are. The network sees addresses. Trust is built from three things that can be checked without personal data.

4.1 Automatic service reputation

The buyer's proxy measures every purchase: whether the answer arrived complete, the inference tokens received against those charged, latency and throughput. Together with public payment events, this produces a service score that anyone can recompute from public inputs and a published formula.

Faults multiply the score down: ghosts (taking a deposit and not serving), repeated overcharging across distinct buyers and circular volume. Measurements are published in aggregate, per session or per day, with anomalies flagged immediately, so a buyer's activity is not exposed purchase by purchase. Manual star ratings are not used.

4.2 The verified-model seal

For open-weight models, auditors recompute random positions of real responses with the public weights and run a statistical test (LOGIC, open source, MIT), plus canary prompts. Offers that pass carry a verified-model seal. The seal proves the model, not the person. Closed models can be sold under the seller's responsibility, but they cannot earn the seal.

4.3 Optional stake

A seller may stake USDC in proportion to its monthly sales. Stake accelerates rewards, unlocks more simultaneous capacity and, later, a fee discount. It is slashed on objective faults or a failed audit, and the slashed amount compensates affected buyers first. Stake never raises the reputation score, so capital cannot buy trust.

Optional public signals such as a verified domain, a .sol name or address age appear as badges, never as requirements.

5. Economics

• Sellers set prices. Gyblos does not price inference.
• One fee for everyone: 3% of each settled payment (initial value), included in the price the buyer sees and split automatically by the channel. Charged from day one.
• Buyers pay only what they use; the rest of the deposit returns at settlement.
• Buyers win on price. Sellers compete, home sellers run on hardware they already own, and token rewards let sellers price lower without paying buyers to trade with themselves.

A managed gateway, for those who prefer an API key and nothing to install, is a separate custodial route at 5–5.5%.

6. The GYB token

6.1 Design

Inference is always paid in USDC. Nobody needs GYB to buy or sell.

6.2 Supply and emission (proposed, to calibrate)

• Maximum supply: 1,000,000,000 GYB, with a maximum schedule that halves every two years.
• Emission follows usage: each epoch, sellers receive min(schedule, R × verified weighted USD), with R = 50 GYB per dollar. Anything not emitted stays in the reserve.
• Buyer weight: a buyer wallet counts for nothing until it has spent with at least 3 distinct sellers over at least 2 epochs.

6.3 Why it resists farming

We modeled a network of 40 home sellers, 15 cloud sellers and a 20-node farm with fake buyers (the model ships with this paper's sources).

• With a fixed schedule, an honest home seller would earn about 8× its sales and farming would pay at a valuation of under US$1.5M. Tying emission to verified usage removes that surplus.
• Without KYC, per-owner caps do not work, because a farm just splits into more wallets. Weighting buyers does: fresh sybil wallets weigh zero and the farm earns nothing.
• A farm that "seasons" its buyers by spending real money with honest sellers only breaks even above a ~US$21M valuation, and in doing so it creates real demand for the network.
• Locked rewards act as stake that sellers never had to deposit: a cheater detected in the first month forfeits about 80% of what it earned.

The model's inputs are assumptions and will be recalibrated with pilot data.

6.4 Audits

Audits are paid from the 10% audit bucket, in three stages: first Gyblos audits openly (its audit rewards go to the treasury, not the team); then sellers audit each other, randomly assigned among nodes serving the same open-weight model and never between nodes of the same owner, with disputed results re-audited; finally, open auditors with stake.

7. Giving intelligence away

A market should also let intelligence be given. Planned mechanisms, to be designed with the community:

• Sponsored channels. Anyone can fund a payment channel for someone else, such as a student, a school or a nonprofit, with a ceiling and an expiry. The recipient uses inference without holding crypto.
• Free offers. A seller can publish capacity at price zero for chosen recipients or programs. Served requests still count toward the seller's service reputation.
• Access fund. Part of the ecosystem bucket and voluntary donations fund access in regions and communities without it.

The hard part is preventing free capacity from being captured by bots. Sponsors choose recipients, budgets are capped and access programs are run by people who know the communities they serve.

8. Status and roadmap

The inference gate, including the review corrections, passed a paid five-case devnet regression on Oct 18, 2026 with a test SPL token: three honest calls, detected within-ceiling overcharge and an interrupted engine with zero charge/full refund. All ten deposit/settlement transactions were independently checked as finalized; buyer SOL was unchanged. The overcharge remained paid: detection does not prevent or compensate fraud. Another 46 private local tests cover admission, buyer validation, observation privacy, transaction visibility and measurement integrity. Signed seller receipts, public reputation metrics, model authenticity, overcharge protection and seller-process crash recovery remain planned; local detection does not establish content correctness.

9. Risks and open questions

• Measurement integrity. Buyer measurements are self-reported until sellers sign receipts; until then they count only when a pattern appears across many buyers.
• Audit privacy. Re-executing responses requires sharing some context with an auditor; who receives it and what is kept will be published before launch.
• Token parameters are proposals from a model and will change with real data.
• Regulation. Token and payment rules differ by country; a non-transferable reward token reduces but does not remove that risk.
• Upstream terms. Reselling closed-model subscriptions often breaks provider terms; sellers answer for what they connect.
• Future public sale. Compatibility of day-one locked rewards with a later public sale (for example on MetaDAO) is still to be confirmed.

Sources

• LOGIC, verification of open-weight inference: https://github.com/context-labs/logic
• x402 upto scheme on Solana: https://github.com/x402-foundation/x402/blob/main/specs/schemes/upto/scheme_upto_svm.md
• 8004 agent registry on Solana: https://github.com/QuantuLabs/8004-solana
• Internal design records: Gyblos product specification (GYB-DOC-005), token decision (GYB-DOC-009) and token model. They will be published with the open-source repositories.