Run AI on your machine.
No cloud required.

Byte runs local models like Llama and Mistral for free, right on your machine. Prefer cloud power? Bring your own key for GPT-5, Claude, or Gemini too.

See how it works

Open source · No tracking · Your data stays on your device

Byte App
macOS Windows Linux

Every model.
One app.

Local models like Llama and Mistral run completely free, right on your machine. Want cloud power instead? Bring your own key for GPT-5, Claude, Gemini, and more.

See how it works
OpenAI
OpenAI
Claude
Claude
Gemini
Gemini
xAI
xAI
+12 more
Byte Demo

Built different.

Most AI apps charge a monthly fee just to talk to models. Byte runs local models free, or you can bring your own key. No markup, no middleman.

Featured

16 Providers

OpenAI, Claude, Gemini, Mistral, Groq, HuggingFace, Together AI, and more.

Get Byte for freeCompare pricing

See it in 60 seconds.

Free.
Not freemium.

Byte is open source and free to download. Local models cost nothing to run. If you use a cloud provider, you pay them directly at their published rates, with no markup from us.

$0
to Byte, always
  • All 30+ models supported
  • Unlimited conversations
  • Full model control
  • Built-in memory
  • Open source

What's next.

Byte ships fast and stays open about where it's headed. Here's what we're building right now.

In progress 3
Windows app
Queue message
Voice input
Planned 1
Linux app

Answered briefly.

Yes, Byte is completely free, always. Local models cost nothing to run. If you use a cloud provider's API, you pay them directly at cost, with no markup from us.
Yes. Byte works with Ollama and LM Studio, so you can run models like Llama and Mistral entirely on your machine. No API key, no internet connection, and no cost.
You paste your existing API key from any provider (OpenAI, Claude, Gemini, etc.) into Byte. Your key stays securely on your device and is never sent anywhere. You pay providers directly at their published rates.
Byte supports 16 AI providers with 30+ models: OpenAI (GPT-4, GPT-5), Claude (3.5, 4), Gemini (2.0, 2.5), Mistral, Groq, Perplexity, OpenRouter, HuggingFace, Together AI, Replicate, Cerebras, Ollama, LM Studio, and more. Plus web search.
Yes. Byte is open source and runs entirely on your device. Your API keys and conversations never leave your computer, with no tracking, no telemetry, and no cloud dependency.
Byte runs on macOS, Windows, and Linux, with the same app and the same features on every platform.
Yes. Byte lets you switch between different AI models instantly, even in the middle of a conversation. Compare outputs from GPT-5, Claude, and Gemini side by side without leaving the app.
Available Now

Download Byte.

Free, open source, and ready in under a minute.