01
Instant Compatibility
Know in seconds if your hardware can run any LLM.
Readiness92%
Local AI benchmark
Can your machine run local AI?
Discover which LLMs your computer can run locally, how fast they'll be, how much RAM/VRAM and disk space they need, and how to install them.
Open-weight models
0+
Curated, normalized catalog.
Families tracked
0
Diverse model families.
Backends supported
0
Ollama, MLX, vLLM and more.
Routes surfaced
0
Detect, manual, compare, benchmarks.
Live catalog health
Can this machine run it locally?
0+
models
Fit score
0%
overall local readiness
Coverage
Confidence
Breadth
Source mix
Merged from the latest Ollama, Hugging Face and fallback catalog sources
1484 entries
Ollama225 / 15%
Hugging Face1207 / 81%
GPT4All32 / 2%
Core fallback20 / 1%
Coverage
0%
Confidence
0%
Backends
0
Why CanRunIt?
Know in seconds if your hardware can run any LLM.
02
Speed Estimates
Get realistic token/s predictions for your specific setup.
Readiness86%
03
Easy Install
Step-by-step guides for Ollama, llama.cpp and LM Studio.
Readiness78%
04
No Cloud Needed
100% local, 100% private. Your data never leaves your machine.
Readiness88%
What can I run?
Pick a machine, choose a use case, get a real recommendation.
This stays practical: no vague promises, just a ranked catalog slice with fit scores, speed hints, and quick jump points to the full app.
Selected profile
RTX 4060 laptop
Smooth for 7B, usable for many 14B Q4 models.
RAM
16GB
VRAM
8GB
Disk
150GB
Top matches
RTX 4060 laptop + Chat
0%
best fit
Best option for your current profile
Loading catalog
Explore the app
The pages you already have should be visible from the home screen, not hidden in the dark.
Detect
Scan your hardware automatically.
Auto
Manual
Enter RAM, VRAM, and storage by hand.
Precise
Use case
Match models to chat, code, or agents.
Intent
Catalog
Browse the full normalized model base.
1484
Compare
Put models and machines side by side.
Rank
Hardware
Known GPUs and machine presets.
GPU
Benchmarks
Community-style speed reports.
tok/s
Donate
Support the project without a paywall.
Open
Popular Models
Most benchmarked models on CanRunIt this week
Powered by
The stack stays open, fast, and practical. Each partner below is a live source, runtime, or research anchor.
Ready to find your perfect local AI?
Takes less than 2 minutes. No sign-up required.