Docs

TamgaBase Documentation

Complete documentation for TamgaBase API and distributed compute infrastructure.

API Reference

Complete REST API documentation with examples in cURL, Python, JavaScript, and Go.

SDKs & Libraries

Official SDKs for Python, Node.js, Go, Rust, and Java. Type-safe and well-documented.

CLI Guide

Command-line interface for power users. Deploy, manage, and monitor from your terminal.

Containers Guide

Deploy and manage Docker containers on our distributed network. Zero configuration required.

Kubernetes Integration

Connect your K8s cluster or create a managed one. Native Kubernetes API compatibility.

Tutorials

Step-by-step guides for ML training, rendering, game hosting, and scientific computing.
API First

Everything is API

Our REST API, SDKs, and CLI give you complete control over your compute infrastructure. Integrate in minutes.

Python SDK 
import tamga_run as tr

# Initialize client
client = tr.Client(api_key="tr_sk_...")

# Deploy a GPU instance
instance = client.instances.create(
    gpu="RTX 4090",
    image="pytorch:latest",
    duration=3600,
    region="eu-west"
)

print(f"Instance {instance.id} running")
CLI 
# Install the CLI
$ npm install -g @tamga/cli

# Authenticate
$ tamga login

# Deploy a container
$ tamga deploy \
  --gpu RTX4090 \
  --image pytorch:latest \
  --duration 3600 \
  --region eu-west

# List instances
$ tamga ps
# → 3 running, 0 stopped
FAQ

Frequently asked questions

Everything you need to know about Tamga Run.

Tamga Run is a distributed compute network. We connect compute resource providers with users who need GPU, CPU, and container capacity. Think of it as a platform for sharing unused hardware and renting compute power.

You only pay for the compute you use, billed by the second. GPU pricing ranges from $0.18/hr for RTX 3060s to $3.49/hr for H100s. CPU instances start at $0.03/hr per vCPU. There are no setup fees, and you can stop any instance at any time.

Yes! Share your unused GPU, CPU, or storage and earn passive income. Our platform supports hardware providers in monetizing idle resources.

We support a wide range of hardware: NVIDIA GPUs (RTX 3060 through H100), AMD GPUs, Intel and AMD CPUs, Apple Silicon (M-series), and any Docker-compatible system. Our platform automatically benchmarks and verifies all hardware before it goes online.

Absolutely. You can run any Docker container on our network. We also offer native Kubernetes integration — connect your existing K8s clusters or spin up new ones with a single click. Our platform automatically handles node discovery, scheduling, and scaling.

Security is our top priority. All workloads run in isolated containers with hardware-level GPU partitioning. Data is encrypted in transit and at rest. We support VPC peering, private networking, and regular security audits.

Ready to join the compute marketplace?

Download TamgaControlCenter and start earning, or rent hardware and start building.

Download for Free → Rent Compute