running on your device

AI that actually
runs locally.

Pallas is a private AI assistant that runs on your machine. No subscription, no cloud, no data leaving your computer.

Download for Windows Use in browser
Pallas — Local AI
Explain quantum computing
Write a cover letter
Debug my Python code
What's the weather today?
Can you explain what a neural network is, in simple terms?
Pallas 1.1
Think of a neural network like a bunch of tiny decision-makers chained together. Each one looks at the input, makes a small judgment, and passes it along. After training on millions of examples, the chain gets really good at recognizing patterns — like faces, words, or even emotions.

Everything runs
on your computer.

No data sent to the cloud. No monthly fee. Just a fast, private AI that's always available — even offline.

🔒

Completely private

Your conversations never leave your machine. No telemetry, no training on your data, no corporate servers involved.

Genuinely fast

Responses in seconds, not minutes. Running locally means no network lag, no queue, no rate limits.

🌐

Optional web search

When you need current info, Pallas can search DuckDuckGo and pull in live results — no API key required.

💾

Chat history

Create an account to save and revisit past conversations. Everything stored locally on your device.

🤖

Multiple models

Switch between Pallas 1 (fast 3B), Pallas 1.1 (smart 7B), and Pallas 1.2 Max (powerful) any time.

🆓

100% free

No subscription, no credits, no paywalls. Open source and free forever. Just download and run.

Pick your speed.

Three models that run entirely on your hardware. No GPU required.

Fast

Pallas 1

3B parameters · Llama 3.2

Response time~2s
RAM usage~2 GB
Best forQuick answers
Recommended

Pallas 1.1

7B parameters · Qwen 2.5

Response time~5s
RAM usage~5 GB
Best forMost tasks
Max

Pallas 1.2 Max

7B parameters · Qwen 2.5

Response time~5s
RAM usage~5 GB
Best forLong, detailed tasks

Your AI, your machine.

Download the Windows app or open it in your browser — free, forever.