Decentralized AI Compute

Compute Without Limits.

Access powerful GPUs, deploy AI models, and run intelligent applications through one decentralized compute infrastructure.

Platform catalog at a glance. Live capacity and network metrics appear here once production regions come online.

Platform

Everything you need to run AI workloads.

From a single GPU for fine-tuning to autoscaling inference endpoints — rent, deploy and pay from one console, with every compute minute backed by a verifiable receipt.

GPU Marketplace

Compare H200, H100, A100, L40S, RTX 5090 and more by VRAM, bandwidth and hourly rate. Launch in a few clicks.

Browse GPUs

Serverless AI

Deploy Hugging Face models or your own Docker container as an autoscaling endpoint. Scale to zero when idle.

Deploy a model

AI Playground

Try image, video, text and voice models side by side. Tune parameters and export what you create.

Open Playground

Developer API

One REST API for instances, inference and billing. Scoped keys with daily spending caps.

Get an API key

Transparent Billing

Per-second metering and prepaid credits. See exactly what each GPU, model and request costs.

View billing

On-chain Receipts

Usage receipts are hashed into hourly Merkle roots so anyone can verify what was billed.

See receipts
How it works

From wallet to workload in five steps.

01

Connect your wallet

Sign in with a supported wallet. Your address is your account — no passwords.

02

Add ecosystem credits

Deposit a supported token and receive prepaid compute credits.

03

Choose a GPU or model

Rent a dedicated GPU instance or pick a model for serverless inference.

04

Run your workload

SSH into your machine, call your endpoint, or generate in the Playground.

05

Track and verify

Monitor spend in real time and verify every receipt against its Merkle root.

GPU Marketplace

Data-center and workstation GPUs, priced by the hour.

Transparent list rates with no long-term commitments. Stop an instance and billing stops with it.

View all GPUs
Developers

Built for engineers who ship.

A clean REST API at . Authenticate with a scoped key and call inference, manage instances and pull usage data from your own stack.

  • Scoped keys with daily spending limits and instant revoke.
  • OpenAI-style inference payloads for fast migration.
  • Usage endpoints that match the receipts you can verify.

        
Networks

Settlement networks

Credits, payments and receipt anchoring run on-chain. Connect with the wallet you already use.

PETEN Ecosystem
$PETEN
PETENTIA

PETEN is the ecosystem token of PETENTIA — designed to connect the people who need compute with the infrastructure that provides it.

Compute payments

A native way to pay for GPU time, inference and API usage across the platform.

Ecosystem utility

Aligns users, builders and GPU providers around one shared infrastructure.

Compute infrastructure vision

An open market for AI compute where pricing and usage are transparent and verifiable.

Start building on PETENTIA.

Open the console, explore the GPU catalog and deploy your first model in the preview environment.