VRAM Calculator

Find out what AI models your hardware can actually run.

This tool is coming soon. For now, here’s the manual version:

The Math

VRAM needed = (model parameters × quantization bits) ÷ 8 + overhead

QuantizationBits per parameterUse case
Q4_K_M4Best quality/size ratio — recommended
Q5_K_M5Higher quality, more VRAM
Q8_08Near-original quality, 2x size
F1616Full precision, rarely needed locally

Quick Reference Table

Model SizeQ4_K_M VRAMQ8 VRAMMin RAM (no GPU)
3B2.5 GB4 GB8 GB
7B5 GB8 GB16 GB
13B8 GB14 GB32 GB
30B18 GB32 GB64 GB
70B42 GB75 GB128 GB

GPU Tier Guide

GPUVRAMBest Models (Q4)
RTX 3060 / 40608 GB3B, 7B
RTX 3080 / 407010-12 GB7B, 9B
RTX 3090 / 409024 GB13B, 30B (tight)
2× RTX 309048 GB30B, 70B (tight)

Coming Soon: Interactive Calculator

We’re building a free tool that will:

  • Input your hardware (GPU model, VRAM, RAM)
  • Output compatible models with expected performance
  • Show optimal settings (quantization, context length, batch size)
  • Generate deployment scripts for your exact setup

Paid tier ($29/mo) unlocks:

  • Full Docker Compose scripts for your hardware
  • Monitoring dashboard templates (Grafana + Prometheus)
  • Pre-built n8n AI workflow templates
  • Priority tutorial requests

Get notified when it launches →