Providers

Turn your idle GPU into income.

Scalattice puts the GPU you already own to work on paid inference jobs. Install one command, keep the majority of every completed job.

Start earning Estimate earnings
  • Majority revenue share
  • $0 to connect
  • Pause anytime

Agent

One command per machine.

Add a GPU on Scalattice Cloud, install the agent with your machine token, and manage everything from the dashboard. Agent source is on GitHub for transparency.

# 1. Add a GPU at scalattice.cloud/providers
# 2. Install on your GPU machine (token required)
curl -fsSL https://scalattice.cloud/install/agent \\
  | sh -s -- --token slt_provider_…

scalattice-agent status
# → Scalattice Cloud: connected
Typical setup in under five minutes, your machine may require updated drivers
Any GPU Consumer or datacenter
Per job Paid when inference runs
Live Earnings on Scalattice Cloud
Uptime Offer nights, weekends, or 24/7 on your schedule.
Payouts Track earnings live on Scalattice Cloud.
Control GPUs, models, schedules, and hardware control in Scalattice Cloud.
Transparency scalattice-agent

Estimate

What could your GPU earn?

We support any device.
Adjust the calculator to fit your earning & specs.

Rates and VRAM requirements are live from the Scalattice catalog. Estimates use published output-token rates and the majority of developer token spend on completed jobs, modelled at Q4 quantisation with throughput scaled by model size. Selected models must fit at least one GPU class you set. Actual earnings vary with live demand, model mix, GPU throughput, quantisation, batching, and uptime, and are not guaranteed. Figures are illustrative only.

$0 Estimated monthly
Estimating… $0/hr Setup $0 Min payout $10
Start earning

Onboarding

Three steps to your first payout.

Sign in, add each GPU on Scalattice Cloud, install the agent with its machine token, and manage capacity from the dashboard.

Step 1

Register

Sign in at Scalattice Cloud, open Providers, register, and add a GPU to get a machine token.

Step 2

Install

Run the install command on a Windows or Linux machine with --token Check the open source agent

Step 3

Earn

Enable models, set your schedule, and track payouts on Scalattice Cloud.

# Token from scalattice.cloud/providers
curl -fsSL https://scalattice.cloud/install/agent \\
  | sh -s -- --token slt_provider_…

# Or set the token later:
scalattice-agent set-token --token slt_provider_…
scalattice-agent status
One agent per GPU machine

Hardware

Built for real machines.

Consumer and datacenter GPUs on Windows and Linux. NVIDIA recommended for the smoothest setup.

NVIDIA

CUDA-ready

From RTX workstations to rack-scale datacenter GPUs. Turn the CUDA hardware you already own into earning capacity.

Schedule

Control your earning

Choose when each machine is available, or pause earning when you need it back.

GPU hardware eligible for the provider network
RTX, workstation, and rack GPUs
$0 connection fee No cost or fees to connect compute as a provider.
Pause anytime Turn off a machine's schedule from Scalattice Cloud.

GPU and driver licence compliance is your responsibility - Terms of service

Safety

Inference-only jobs.

Customer workloads are inference-only. You control when each machine earns, and can pause from the dashboard.

Inference-only jobs

Inference runs locally inside the agent on your hardware.

Pause earning

Set a machine's schedule to Turned off in Scalattice Cloud when you need it back.

Metered jobs

Completed jobs are logged with model, tokens, and payout in your provider dashboard.

How the network routes jobs · Open-source agent

Platform

Inference hosting on the Scalattice network.

Connect your GPU once. Inference jobs find your hardware when you're available - you keep the majority of developer spend.

Inference-only catalog models

Enable curated models from the dashboard. Weights are mirrored by Scalattice.

Majority token revenue share

Earn the majority of developer token spend on completed inference jobs.

$0 connection fee

Connect machines at no cost. On-demand payouts from $10.

Demand ratings

Per-model demand rating (1 to 5) shows where jobs are most likely to route.

Schedule control

Always on, paused, or UTC hour windows from Scalattice Cloud.

Open-source agent

MIT-licensed agent for Linux and Windows. Audit what runs on your machine.