Quick Keypoints
- Detects local GPU, CPU, and VRAM specifications via browser WebGL/WebGPU.
- Assigns runnability scores to various open-source models based on specs.
- Estimates generation speed (tokens/sec) and memory footprint for quantization levels.
- Privacy-focused diagnostic runs entirely client-side without data leaving the device.
What is CanIRun.ai?
Local AI model hardware compatibility and runnability checker.
CanIRun.ai is a privacy-first hardware checking tool for local AI execution. It runs fully in your web browser, calling WebGL and WebGPU APIs to inspect your system specs. It compares your CPU, RAM, and GPU VRAM capacity against a database of open-source models to report exactly which LLMs can run on your device.
Who Needs CanIRun.ai?
AI developers, open-source enthusiasts, and users wanting to run models locally before downloading files.
Primary Use Cases
- Detecting local CPU, GPU, and VRAM specifications without downloading software.
- Scoring runnability compatibility for open-source LLMs (Google, Meta, Mistral) on local machines.
- Estimating local generation speed (tokens/sec) and memory requirements for different GGUF quantization levels.
Important Features
- Client Hardware Scan: Detects system capabilities locally via web APIs in seconds.
- Runnability Gauge: Labels compatibility ranges from 'Runs great' to 'Too heavy' for each model.
- Quantization Chart: Details VRAM requirements for different precision levels (Q2 to F16).
Current Updates About CanIRun.ai
CanIRun.ai recently updated its index to include DeepSeek-V3 and Google's Gemma 2 series models.
Editorial Rating
4.7 / 5.0
Pricing Plans
| Plan | Price |
|---|---|
| Free ToolWebGL hardware diagnostics, model capability scoring, and quantization details | $0 |