✨ Perceptron is now in Public Beta ✨
SuperIntelligence At Your Fingertips
World's first Crypto-native NeoCloud

Access frontier AI models, Agents, and powerful tools for your AI workflows, all without creating an account or submitting an email, as all services are billed in USDT. Instantly Login with a wallet like MetaMask, Phantom, Trust, and Base wallet.

Install

Personas

Perceptron Personas are AI Agents with personality, memory, and advanced skills. They remember your conversations, work on all kinds of tasks, and live persistently on Perceptron Cloud, without requiring any setup or local installation.

Memory
Voice
Avatars
Subagents
Jensen avatar Nova avatar Vega avatar Scarlet avatar Luna avatar Lisa avatar Eliza avatar Marie avatar Turing avatar Kim avatar

Multiple Avatars

Jensen
Nova
Vega
Scarlet
Luna
Lisa
Eliza
Marie
Turing
Kim
You
Please build a personal blog for my traveling adventures, and publish it with Perceptron Domains
I'll get started building your personal blog, and i'll publish it on myadventures.pct.site once it's done.
Powered by Perceptron Neo harness
Login → Telegram Bot →

Frontier AI Models

Deimos

Our flagship AI Model for frontier Agentic Intelligence

Deimos
Deimos 3.5 Air

Fast and affordable, frontier quality

Context 1M
Modalities Text
Input $0.5 / 1M
Output $1 / 1M
Deimos
Deimos 3.5 Pro

Balanced capabilities for everyday work

Context 1M
Modalities Text
Input $2 / 1M
Output $6 / 1M
Deimos
Deimos 3.5 Ultra

Our most capable model for your toughest tasks

Context 1M
Modalities Text, Image
Input $5 / 1M
Output $25 / 1M

Featured Models

Kimi K3
Context 1M
Input Modalities Text, Image, Video
Output Modalities Text
moonshotai/kimi-k3
GLM 5.3
Context 1M
Input Modalities Text
Output Modalities Text
z-ai/glm-5.3
Deepseek V4 Pro
Context 1M
Input Modalities Text
Output Modalities Text
deepseek/deepseek-v4-pro
Seedance 2.5
Resolutions 480p, 720p
Input Modalities Text, Image
Output Modalities Video
bytedance/seedance-2.5
Seedream 5.0 Pro
Resolutions 1K, 2K
Input Modalities Text, Image
Output Modalities Image
bytedance/seedream-5.0-pro
Kokoro 82M
Context 4K
Input Modalities Text
Output Modalities Audio
hexgrad/kokoro-82m

One Unified API

curl https://perceptron.cloud/api/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $PERCEPTRON_API_KEY" \
  -d '{
  "model": "bartmoss/deimos-3.5-pro",
  "messages": [
    {
      "role": "user",
      "content": "What is the capital of Japan?"
    }
  ],
  "reasoning": {
    "enabled": true
  }
}'
Explore Models →

Perceptron Studio

Create Videos, Images, Animations, and Sound with frontier generative AI Models. Every creation ready to download, remix, and share.

Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example Studio image generation example
Video
Images
Animation
Sound
Open Studio →

Agentic APIs

Connect existing AI Agents like Hermes and OpenClaw to the Perceptron API, and enhance their capabilities with fast web search, domains, and data feeds

Domains

Instant domain names that help Agents host their web applications and websites on the internet

Boxes

Local and Cloud Virtual Machines (VMs) and SandBoxes usable by personas for all kinds of computer-use tasks

News Feed

Real-time news streams to inform Agents of market moves, natural disasters, tech news, and global politics

API Docs →

Local Models & Intelligence Mining

Run AI models on your own compute hardware and clusters. Use them for your own Perceptron usage, or sell your spare compute to other Perceptron users and earn rewards in USDT.

Your Compute, Free Intelligence

Run AI models locally with the Perceptron app, and route your chat, Persona, and API requests to your own nodes. Your own hardware serves you free, with optional automatic cloud fallback when your node is offline.

Earn Passive Income

Sell your spare compute to other users on Perceptron. Earn rewards on every request you serve, level up your reputation as an independent inference provider, and withdraw your earnings as USDT anytime.

Mining Nodes are your devices (and connected XPUs) running the Perceptron native app available for macOS, Linux, and Android.

Start Mining →

Perceptron TUI

Manage services, API Keys, and track your usage with a fast, keyboard-driven Terminal User Interface (TUI) with vim key bindings and customizable themes

Blazingly Fast: Written in Rust for smooth rendering
Keyboard Driven: Navigate everything without touching your mouse
Built-in EVM Wallet: Buy credits without leaving the terminal, no credit cards needed
Real-time Metrics: Track your token usage in real-time
Customizable Themes: theme picker and editor included
Perceptron TUI Interface

Autonomous Installation

Hermes/OpenClaw Agent prompt:

Use the Perceptron Skill at https://perceptron.cloud/api/SKILL.md to install the Perceptron Terminal UI,
then use the TUI to generate a wallet, and create a new client API key

Manual Installation

curl -fsSL https://perceptron.cloud/install.sh | bash

Powered by leading Blockchain Networks

FAQ

+ What is Perceptron?
Perceptron is the Solar System's first Crypto-native NeoCloud. It provides frontier AI Models, Agents, and powerful tools for your AI workflows.
+ Do I need to create an account?
Perceptron allows you to use all features and services without creating an account or submitting an email, as all services are billed in USDT, a crypto Stablecoin. All you need to sign in is a crypto wallet, and your wallet address is your username.
+ How does billing work?
Purchase Perceptron Credits with USDT on any of the supported Blockchain Networks including Base, HyperEVM, Ethereum, Binance Chain, Polygon, Arbitrum, and Monad. All services are billed from your purchased credits
+ Can I use Perceptron with any AI Agent?
Yes. The Perceptron API works with any Agent harness or framework. Perceptron Neo harness, Nous Hermes Agent, OpenClaw, OpenCode, pi, Google Agent Executor, custom agents, or any agent system with the ability to make HTTP API calls.

The Agentic Cloud