NeuralPool Dashboard
LIVE
Your compute + network overview · auto-refreshes every 15s
$0.00
Credit Balance
$0.00
Total Spent
0
Running Instances
0h
GPU Hours Used
No instances running. Rent a GPU →
Network Monitor
—
Active Miners
—
Block Height
—
AI Jobs Done
$—
USDC Paid Out
—
NPOOL/day
GPU Availability
My Usage (last 7 days)
No usage yet.
Rent GPU
Pay by the second — cancel any time. 25% off when paying with $NEURAL.
API Keys
Use the OpenAI-compatible inference API at https://api.neuralpool.io/v1
Quick Start
$ pip install openai
from openai import OpenAI
client = OpenAI(
api_key="npool_...",
base_url="https://api.neuralpool.io/v1"
)
resp = client.chat.completions.create(
model="llama3:8b",
messages=[{"role":"user","content":"Hello!"}]
)
print(resp.choices[0].message.content)
Available Models
| Model | Context | Input | Output |
|---|---|---|---|
| llama3:8b | 8 192 | $0.10/1M | $0.10/1M |
| llama3:70b | 8 192 | $0.80/1M | $0.80/1M |
| mistral:7b | 32 768 | $0.08/1M | $0.08/1M |
| mixtral:8x7b | 32 768 | $0.45/1M | $0.45/1M |
| codellama:34b | 16 384 | $0.55/1M | $0.55/1M |
| phi3:mini | 4 096 | $0.04/1M | $0.04/1M |
Billing
Prepaid credits — unused balance never expires
$0.00
USDC Balance
0
$NEURAL Balance
$0.00
All-Time Spend
Add Credits (USDC)
$10
$25
+2% bonus
$100
+5% bonus
$500
+10% bonus
$NEURAL discount: Pay with $NEURAL and get 25% more compute. Send $NEURAL to your deposit address below.
Loading…
Transaction History
No transactions yet.
Documentation
Quickstart guides and API reference
</>
Inference API
OpenAI-compatible. Drop-in replacement for GPT-4.
⚡
GPU Rental
SSH access to raw NVIDIA compute by the second.
⊞
REST API
Manage instances and billing programmatically.
SSH Access Guide
# After launching a GPU instance, you get SSH credentials:
$ ssh root@203.0.113.42 -p 22 -i ~/.ssh/neuralpool_key
# NVIDIA drivers + CUDA pre-installed. Check GPU:
$ nvidia-smi
Thu May 6 00:00:00 2026
+-------------+ NVIDIA RTX 4090 24 GB GDDR6X
GPU utilization: 0% | Memory: 0 / 24576 MiB
Billing Model
| Detail | Value |
|---|---|
| Billing granularity | Per second |
| Minimum session | 60 seconds |
| Payment | Prepaid credits (USDC or $NEURAL) |
| $NEURAL discount | 25% more compute per dollar |
| Idle instances | Billed at 50% rate (GPU reserved) |
| Data egress | Free (first 100 GB/month) |
Download Daemon
Install the NeuralPool daemon to start mining $NEURAL and serving AI jobs with your GPU.
Detected: Detecting… — recommended download highlighted below.
🐧
Linux (x86_64)
Ubuntu 20.04+ · Debian · CentOS 8+
NVIDIA driver 525+ required
↓ neuralpoold (linux/amd64)
↓ ARM64 (Raspberry Pi / Jetson)
macOS
Apple Silicon (M1/M2/M3/M4)
No NVIDIA GPU required — runs validator node mode
↓ neuralpoold (darwin/arm64)
↓ Intel Mac (amd64)
⊞
Windows
Windows 10/11 x64
Installs as Windows Service via PowerShell
↓ neuralpoold.exe
PS> irm https://neuralpool.io/install.ps1 | iex
◈
Validator Node (any PC)
No GPU required · earns USDC yield
Verifies epoch payouts, signs batches, collects 1/3 of job pool fees
↓ Validator (Linux)
↓ Validator (macOS)
One-liner install (Linux/macOS)
$ curl -sSL https://neuralpool.io/install | bash -s -- --id=YOUR_PROVIDER_ID
Don't have a provider ID yet? Sign up as a miner →
System Requirements
| Component | Minimum | Recommended |
|---|---|---|
| GPU | Any NVIDIA (PoW mode) | RTX 4090 24 GB (AI + PoW) |
| NVIDIA driver | 525+ | Latest stable |
| RAM | 4 GB | 32 GB |
| Storage | 20 GB free | 100 GB SSD |
| Network | 10 Mbps up/down | 1 Gbps symmetric |
| OS | Ubuntu 20.04 / Win10 / macOS 12 | Ubuntu 22.04 LTS |
| Ollama | Only for AI jobs (24 GB+ VRAM) | Auto-installed by script |
Release History
| Version | Date | Notes |
|---|---|---|
| v0.1.0 | 2026-06-01 | Genesis release — PoW mining, AI inference, validator node, auto-updater |