About this Local LLM Hardware Calculator
This interactive tool helps developers and AI enthusiasts determine the optimal Large Language Model to run locally on their personal hardware. By selecting the platform and available memory, the tool calculates the maximum parameter size for GGUF quantized models that fit in memory without swapping. It recommends modern models like DeepSeek-R1, Qwen2.5-Coder, Phi-4, Llama 3.3, Gemma 3, GPT-OSS, and Devstral, providing exact Ollama CLI commands, and lets users search a catalog of over 75 open source models directly.
skiptodone:~$ check-compatibility –machine
Can my machine run this LLM?
Tell it your RAM, VRAM, and platform. It instantly calculates the best local models that fit your hardware, or search over 75 open source models to check any specific one.
01 – Your hardware
02 – Recommendation
Sizes are estimates based on standard Q4 GGUF quantization weights plus a working context buffer. Actual usage varies with context length. Always leave OS headroom.