Strata: run a giant AI model on your own gaming PC for free

Strata is a free, open-source tool that runs Qwen3.8-Flash-Next, a 125-billion-parameter AI model, on an ordinary gaming PC. It chats, writes code and reads pictures, and nothing leaves your computer.

Set it up

1. Check your PC

   You need an NVIDIA RTX 20/30/40/50 series or a supported AMD Radeon card with 12 GB of VRAM or more, 32 GB of RAM or more, about 80 GB of free disk (an SSD helps), and Windows 10/11 or Linux.

2. Download Strata

   Download the ZIP from GitHub and unzip it, or clone the repository.

   Copygit clone https://github.com/Niko1221/Strata

3. Start the installer

   On Windows double-click START-HERE.bat. On Linux run ./setup.sh in the Strata folder. Press Enter at each question to take the recommended answer.

   Copy./setup.sh

4. Wait for the model

   It downloads the model (about 70 GB) and starts it. If the download stops, run the installer again and it continues where it left off.

5. Open the app

   Your browser opens the Strata app. Next time, run the same file again: it starts right away.

   Copyhttp://127.0.0.1:8080

Good to know

• While the model starts, your PC can be slow or stop responding for 1-3 minutes. The README says this is normal.
• Your RAM decides which model size fits: 32 GB runs the Coder version, 64 GB runs every size.
• The 94 tokens per second figure is the project's own measurement on an RTX 5070 with the smallest size (Q2_0).
• It also offers an OpenAI/Anthropic-style API on localhost for your apps and coding agents.
• Facts are from the README on 7 October 2026.

Published 2026-10-06.