Free demo · No account needed

Try a real AI.
Right now.

We spin up an actual server in a real datacenter, install the Inference-X engine, load an AI model, and give you a live chat interface. Everything is erased when you're done. No data stored. No account created.

✓ No account · ✓ No credit card · ✓ 30 min · ✓ Auto-erased

ix-demo · waiting...

🌍

00:00Inference-X Demo Terminal —

🐙

Inference-X Community

git.inference-x.com — Open source · Community-built · BSL License

📦 16 repos

🌍 3 public

👥 Community

Browse All →

Open Source Community

Build the future.
Together.

These three repositories are the foundation. Fork them. Build on them. Propose your changes. The community makes the decisions — no single company controls the roadmap.

⚙ inference-x

elmadani/inference-x

BSL 1.1

Universal AI inference engine — 305 KB, 19 hardware backends, zero dependencies. The core: loads any GGUF model, exposes an OpenAI-compatible API, runs on any hardware from Raspberry Pi to data center GPU clusters.

⭐ —

🍴 —

📝 C++ · CMake

ai-inference c++ openai-compatible offline 19-backends

View Repo → README

🔧 ix-tools

elmadani/ix-tools

MIT

CLI tools, deployment scripts, model downloader, benchmark suite. Everything you need to set up, configure, and monitor your Inference-X deployment. Install scripts for all platforms.

⭐ —

🍴 —

📝 Shell · Python

cli installer benchmarks python

View Repo → README

📖 ElmadaniS

elmadani/ElmadaniS

Public

The founder's public work — mathematical frameworks, philosophical essays, project architecture documents. Understand the vision behind Inference-X: why it was built, where it's going, and the H5→H6 consciousness framework.

⭐ —

🍴 —

📝 Markdown · Math

philosophy mathematics z-equation vision

View Repo → All repos

Want to contribute?

Fork any repo · Open an issue · Submit a PR · Propose your craton · Join the builders

Open Issue Contact →

Configuration Builder

Build your
custom AI.

Configure your IX deployment for your exact hardware and use case. Generate a ready-to-run config file you can deploy anywhere.

🖥 Hardware

🧠 Model

🎭 Persona

📦 Export

Your Operating System

Hardware Backend

Available RAM

8 GB

Models that fit your hardware:

API Port

Context Size (tokens)

Select Model

Custom Model URL

Quantization Preference

Max Tokens per Response

AI Name (optional)

System Prompt

Response Style

Temperature

Generated configuration:

Quick start:
1. Download the config above
2. Download the IX binary: inference-x.com#start
3. Run: ./ix --config ix-config.json

Community Compute

Power the demos.
Earn your place in history.

The Provider Pool.

Every free demo needs real compute. Community providers contribute their idle server capacity. In return, they're credited publicly, gain early access to future IX frameworks, and become part of the infrastructure that democratizes AI.

Pioneer providers will have priority integration when the Echo Relay (federated inference network) launches.

✓Your name/brand displayed as compute provider on every demo

✓Daily cost limit you control — never exceed your budget

✓Keys encrypted server-side — never exposed to public

✓Early access: Echo Relay framework for distributed providers

✓OneCloud, Hetzner, OVH or any API-compatible provider

🔋 Contribute compute

Keys encrypted · Used only for IX demos · Revokable anytime

Active pool contributors

No providers yet — be the first to contribute compute.

Backend	Nodes	Avg tok/s	Load	Status
Loading community nodes...

Pricing — Fair by design

Free for people.
Fair for business.

80% of all commercial revenue flows directly to community contributors. Individuals, researchers, students and small startups pay nothing. Always.

Free Forever

$0/month

For individuals, researchers, students, open-source projects, and startups under $1M revenue.

✓Full IX engine binary

✓All 19 hardware backends

✓All GGUF models

✓Unlimited API calls (local)

✓3 builder configs

✓Community support (Gitea)

—Dedicated cloud instance

Download Free →

Try a real AI.Right now.

The Provider Pool.

🔋 Contribute compute

Try a real AI.
Right now.