Own your AI. Don't rent it.
CanYouRunLLM.com helps you figure out which local AI models you can run on your hardware. One scan, one honest answer.
Why we built this
Renting AI through APIs is convenient, but it's also expensive, rate-limited, and completely dependent on someone else's server. Local AI gives you full control, zero ongoing costs, and complete privacy.
The problem? Figuring out whether your GPU or Mac can run a specific model requires digging through technical specs, calculating memory requirements, and trial-and-error testing. We wanted to make that instant.
So we built CanYouRunLLM.com - a simple tool that tells you exactly which models you can run, at what speeds, with exact setup commands. No guesswork, no spreadsheets.
Our principles
Accuracy
We verify every spec from official sources. Our compatibility calculations are open-source and tested against real benchmarks. If we don't know, we say so.
Privacy
The detector shows you exactly what it will send before upload. No personal info, no tracking pixels, no account required. You can delete your scan anytime.
Open
The detector, compatibility engine, and dataset are all open source. Fork it, verify it, improve it. We'll accept quality contributions.
Fast
No heavy frameworks, no bloat. The site loads instantly and the detector scans in under 10 seconds. Speed matters.
What's next
Right now we're focused on getting the detector app launched and adding measured community benchmarks. After that:
- Expand device coverage (gaming laptops, Ryzen AI, Snapdragon X)
- Add AI harness compatibility (Codex CLI, Claude Code, Cursor)
- Multi-GPU and cluster support
- Mobile device support (iOS, Android on-device LLMs)
- Hardware upgrade advisor