LLM Configurator — Free GPU VRAM Checker for Local AI Models

Name: LLM Configurator — GPU VRAM Checker
Author: LLM Configurator

LLM Configurator is the definitive free tool for checking GPU compatibility with local LLMs. Enter your GPU's VRAM and system RAM to instantly discover which open models you can run — with Ollama install commands, speed estimates, and electricity cost calculations.

GPU & VRAM Hardware Checker — Check your hardware against 95 model families (196 variants)
LLM Model Library — Browse Llama 4, Qwen 3.5, Gemma 4, DeepSeek V4, and more
Local vs Cloud Cost Calculator — Compare local AI vs ChatGPT/Claude API costs
GPU Benchmark Leaderboard — Modelled tokens/sec, fitted to 14 measured llama-bench runs — RTX 4090, M4 Max, and more
Training Dataset Hub — 47 curated fine-tuning datasets
Setup Guides — 113 tutorials from beginner to advanced
Local AI Reports — Biweekly LLM & hardware digest

What fits your GPU

Best LLMs by VRAM tier — every catalogued model that fits 4 GB to 80 GB, sized by total parameters
Best local LLMs for your GPU — ranked for your exact card
VRAM calculator — weights, KV cache and overhead at every quantization
Methodology — the published formulas behind every figure

Featured model families

Llama 4 (Meta) — Mixture-of-Experts: Scout is 109B total / 17B active, Maverick 400B total / 17B active. Memory is sized by the TOTAL.
DeepSeek V4 (DeepSeek)
Gemma 4 (Google)
Qwen 3.5 (Alibaba)
Mistral Small 4 (Mistral AI)

Setup Guides

About LLM Configurator

LLM Configurator is a free, independent tool for the local AI community, created by Jakub Rusinowski, an AI educator and workshop leader on local LLM deployment. Supports 196 open-weight model variants. No account required. No ads. Free forever. Contact: contact@llmconfigurator.com