Open Source · AGPL-3.0

Run AI on
your own machine

Local AI platform for running Llama, Mistral, Gemma and more — entirely on your hardware. No cloud, no subscriptions, no data leaving your machine.

Quick install — Windows · macOS · Linux
uv tool install "git+https://github.com/metiu1/Vortelio#subdirectory=vortelio-pip"
pip install vortelio
No data collection 100% local inference Free forever for local models OpenAI-compatible API
100+
Supported models
€0
Cost for local inference
OpenAI
API compatible
AGPL-3.0
Open source license
Capabilities

Everything AI, on your terms

From local LLMs to image generation and document processing — Vortelio is the complete AI toolkit that runs on your hardware.

LLMs — chat & completions

llama.cpp backend, GGUF format. Run Llama 3, Qwen, Mistral, Gemma, Phi and any model from HuggingFace — entirely on your hardware.

Image generation

Stable Diffusion 1.5 / XL, FLUX.1, Kandinsky via diffusers. Generate images fully offline, no watermarks, no rate limits.

Audio — STT & TTS

Whisper for speech-to-text transcription. Kokoro and Bark for high-quality text-to-speech synthesis. All offline.

Video generation

Text-to-video with WAN 2.1, AnimateDiff and CogVideo-X. Generate video clips locally on your GPU.

3D generation

Image-to-3D and text-to-3D with TripoSR, Shap-E, LGM and TRELLIS. Export meshes directly from your machine.

OpenAI & Ollama API

Drop-in replacement for OpenAI and Ollama endpoints. Works with Cursor, LangChain, LlamaIndex and any compatible tool.

Why Vortelio

Built on trust,
designed for privacy

Fully open source

Every line of code is public and auditable on GitHub. Licensed AGPL-3.0 — no black boxes, no hidden telemetry.

No account needed for local use

Install and run immediately. An account is only required if you want cloud model access.

One command to start

uv tool install vortelio is all you need. No Docker, no complex setup.

Transparent pricing

Local models are free forever. Cloud access is optional, monthly, and cancel-anytime.

$vortelio pull llama3
Pulling llama3:8b… done (4.7 GB)
 
$vortelio run llama3
Starting server on localhost:11434
GPU: RTX 3080 · Layers: 32/32
 
Ready to chat
✓ No data leaves this machine
Get started today

Ready to run AI privately?

Free forever for local models. Upgrade when you need cloud access to frontier models.