wattz
Solana devnetOpenAI-compatibleGroq LPU relay

Power the inference.

Point any OpenAI client at https://api.wattz.fi/v1. Requests route through the gateway to verified capacity; settlement clears on Solana.

substation booting
program idGUDV…TLiU
gateway
relay-live models
The gap

Solana has liquidity, users, and GPUs. It does not have an inference standard.

Every existing piece was solved by someone. Nothing was solved together. Six concrete gaps make AI on Solana feel improvised.

IaaS only
GPU rentals are not inference

io.net and its peers rent bare compute. There is no OpenAI-compatible surface, no model registry, no license enforcement.

Registry gap
No shared model registry

Llama 3, GPT-OSS, Stable Diffusion, and Whisper each ship with distinct licenses. Solana has no canonical way to publish and version them on-chain.

Trust gap
Nodes cannot prove they computed

Renters take GPU logs at face value. Without TEE attestation or ZK proof, a dishonest node can return anything it wants.

Routing gap
No native routing

Which GPU, which region, which model, which price? Every dApp has to invent its own router. There is no protocol layer.

Payment gap
Per-token payments are absent

Inference cost is measured in tokens, not epochs. Without Token-2022 transfer hooks, teams overpay through 30-day invoices.

Fleet gap
Consumer + DC fleets are siloed

Home RTX 4090s and H100 racks should share load. Instead they run on different rails with different economics.

The substation

One protocol. Six busbars. Every AI dApp gets the same wiring.

Wattz is not another GPU rental service. It is the first inference marketplace built to speak the OpenAI dialect Solana already understands.

step 01
Client calls /v1/chat/completions

Existing OpenAI SDKs just swap baseURL. No new client library required. Streaming responses ride the same SSE contract as OpenAI.

step 02
Router picks the transformer

The routing engine matches the request against model availability, region latency, GPU class, and node reputation. Priority slots are staked with $WATTZ.

step 03
Node executes inside a TEE

Selected GPU nodes execute under Intel SGX, AMD SEV, or NVIDIA Confidential Computing. Each response returns an attestation quote plus an optional Risc0 / SP1 receipt.

step 04
Token-2022 streams payment

Per-token cost is settled through a Token-2022 transfer hook, meter by meter. Refunds and disputes settle through the Anchor settlement program (devnet today).

step 05
Model registry stays honest

Every model lives as a PDA with license, checksum, and version. Registry publish scans license text and gates commercial-only models behind KYC.

step 06
Bootstrap nodes keep the grid warm

Bootstrap capacity keeps the grid warm. Today that is a Groq LPU relay; Wattz-operated bare-metal nodes join the pool as they come online.

Four busbars

Every load routes through the same four pieces.

A model registry, verified compute, an honest router, and a streaming payment layer. Nothing else in the middle. Nothing proprietary at the edge.

Model registrydevnet
On-chain source of truth
  • Every model publishes as a PDA with license, weights checksum, and version.
  • The registry crawler scans license text at publish time; commercial-only weights are KYC-gated automatically.
  • License classes cover Meta Llama Community, Apache 2.0, MIT, and OpenRAIL-M.
Catalog5 models, 3 relay-live
LicensesMeta / Apache 2.0 / MIT / OpenRAIL-M
Compute verificationrelay
TEE plus optional ZK
  • Bare-metal GPU nodes run under Intel SGX, AMD SEV, or NVIDIA Confidential Computing, and each response carries an attestation quote.
  • Until the first node registers, inference is relayed and marked kind "relay" with verified:false -- no attestation is invented.
  • Once nodes are live, skeptical clients can demand a Risc0 or SP1 receipt of the tokenizer path.
Todayrelay path, verified:false
On node registerSGX / SEV / NVIDIA CC
Routing enginerelay
Latency-aware and honest
  • The router matches each request against model availability, region latency, GPU class, and per-call price ceilings.
  • Today all traffic routes to the Groq LPU relay; the wire protocol does not change when nodes join.
  • Fallback walks candidates in order, so a single evicted node never breaks a session.
Todayroutes to Groq LPU relay
Cold starttarget: <200 ms
Streaming paymentat TGE
Token-2022 per-token settlement
  • Each settled request splits its fee on-chain: 80% to the node immediately, 10% held as node pending, 5% to the model publisher, 5% project fee.
  • Half of the project fee -- 2.5% of every settled fee -- is burned through a direct SPL Token Burn CPI.
  • Token-2022 streaming settlement activates with the $WATTZ mint at launch; one request settles once, not once per token.
Fee splitNode 80, pending 10, pub 5, proj 5
Burn2.5% of every settled fee
Quickstart

One request. The same wire as OpenAI.

Send this exact call to the live gateway and watch tokens stream back. Swapping an existing OpenAI client is one base URL change.

curl https://api.wattz.fi/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "llama-3.1-8b-instant",
    "messages": [{ "role": "user", "content": "Power the inference." }],
    "stream": true
  }'

Inference is relayed through Groq LPU capacity until the first bare-metal node registers. The wire protocol does not change.

$WATTZ

The meter is the token.

Every inference call is a small transaction. Every token returned is a small settlement. Value accrual is a function of throughput, not narrative.

$WATTZ CAfxcX1xSZ4Uz9pimw97XENdEHwexHrfaRkEkKeHupump
Per-settlement fee routing
settled fee 100%80%node10%node pending5%model publisher5%project fee2.5%burn / SPL CPI
Node stake collateral

Operators post $WATTZ to register a node -- a protocol minimum of 100 $WATTZ, locked for 7 days. Bad attestations and dropped sessions are slashed, and the slashed stake is burned.

Fee burn on every settlement

Each settled inference splits its fee 80/10/5/5 across node, node pending, model publisher, and project. Half of the project fee -- 2.5% of every settled fee -- is burned through a direct SPL Token Burn CPI.

Model publishing

Registering a model to the on-chain registry stakes the publisher share; that publisher then earns 5% of every settlement routed to the model.

Priority routing lane
planned

A stake-weighted low-latency lane is a design intent, not live. Routing today scores price, latency, reputation, and region only.

Community dispute resolution
at TGE

Holder-voted judgments on disputed settlements are planned for TGE. On devnet, disputes resolve inside a one-hour window under admin control.

Distribution
Node operatorsLong-tail rewards + slashing bond
34%
Model hostersPublish Llama, GPT-OSS, and community models
18%
TreasuryBootstrap nodes, grants, security audits
20%
Public saleOn-chain public tranche via Anchor
14%
Team / ContributorsFour-year vesting, one-year cliff
8%
Ecosystem grantsIDE integrations and OSS SDKs
6%
Total supply
1,000,000,000
Emission
Zero fresh mint
Mint auth
Revoked at TGE
Standard
Token-2022
Substation online

Try one inference. Then swap your OpenAI baseURL.

The playground streams live from the gateway at https://api.wattz.fi/v1. Point your existing OpenAI SDK at the same base URL and keep every call you already have.

Inference is relayed through Groq LPU capacity until the first bare-metal node registers. The wire protocol does not change.