quantized.uk

quantized.uk is a VRAM index and hardware calculator that shows which of 122 quantized large language models fit specific graphics cards.

Quantized.uk is a VRAM index and hardware calculator designed for users running large language models locally. It helps users determine which of 122 quantized models fit their specific graphics card, showing exact memory requirements at 4K context, remaining VRAM, and the required run commands. The platform includes tools such as a VRAM calculator, CLI generator, format wizard, and model comparison views. It supports various quant formats like GGUF, AWQ, and EXL2 across hardware including NVIDIA, AMD, and Apple Silicon. Users can filter models by task, hardware, or specific quantization levels to find the best fit for local inference.

quantized.uk at a glance

Best for: Software engineer
Runs on: Web
Pricing: Free
Released: October 2026
Category: AI research tool
Website: quantized.uk

Key features

VRAM calculator for model, quant, and context combinations
CLI command generator for llama.cpp, Ollama, vLLM, and ExLlamaV2
Format wizard for GGUF, AWQ, and EXL2 recommendations
Model comparison tool for side-by-side VRAM and speed metrics
Hardware compatibility index across 122 models
Real-hardware token-per-second speed benchmarks

Use it for

Check if a specific LLM fits on a GPU
Generate run commands for local inference
Compare VRAM and speed for two different models
Find suitable models for an 8GB VRAM card

Alternatives to quantized.uk

llmrun — Check which open models run on your specific GPU or Mac instantly.
RunThisLLM — Provides exact hardware and VRAM requirements for over 295 local AI models.
NanoJev — It runs parallel decisions and outputs probability distributions without output-token decoding.
Every number in a tiny GPT — Inspects every lookup, matrix multiplication, and normalisation in a live transformer.
Vev — Runs locally on NVIDIA hardware with a fast, Jev-compatible API.
Decido — Runs locally on CPU while beating published benchmark scores.
Vegaml — Runs locally on your hardware with calibrated probabilities and conformal answer sets.
SemIf OpenJev — Extracts direct token probabilities without generating text or running JSON repair loops.
Amagine3D — Generates editable CAD source code directly from natural language and images.
Kev — Provides calibrated probabilities and matches the TypeSafe API.
VLMEvalKit — Evaluate large vision-language models across numerous benchmarks with a single command.
Ncm — Generates text significantly faster than diffusion baselines while maintaining high quality.

All quantized.uk alternatives