Runyard is a free hardware-aware AI model browser. You enter your CPU, GPU, and VRAM and it instantly shows every local LLM that will run on your machine, ranked by speed and quality.

How much VRAM do I need to run local LLMs?

8GB of VRAM runs 7B models like Llama 3.1 8B and Mistral 7B at Q4 quantization. 16GB unlocks 13B models. 24GB lets you run Mixtral 8x7B and Llama 3 70B at lower quantization.

What is the best local LLM for my GPU?

Use Runyard at www.runyard.dev — enter your GPU and VRAM and the Model Radar will rank every compatible LLM for your exact hardware, showing estimated tokens per second for each model.

Can I run Llama 3 locally?

Yes. Llama 3.1 8B at Q4 runs on any 8GB VRAM GPU. Llama 3.1 70B needs around 40GB VRAM at Q4, or an Apple Silicon Mac with 64GB+ unified memory.

Model Comparison Matrix | Runyard Tools

P-13

Model size

Quantization

GPU or unified memory

System RAM

Context target

Quality preference

Speed preference

TurboQuant KV compression

Live result

Smaller model wins locally

P-13

The matrix favors the model that preserves fit and responsiveness first, not just theoretical ceiling quality.

7B weight

-3.94 GB

Faster local fit

14B weight

-7.88 GB

Higher ceiling, heavier cost

How It Works

3 inputs. Instant results.

01

Set the scenario

Choose realistic hardware, model, and context assumptions.

02

Read the result

The hero shows a working result instead of a decorative promo block.

03

Act on the outcome

Use the result to adjust fit, speed, quantization, or context.

Features

Everything that powers model comparison matrix.

01

Planning-first

Built to make local-AI decisions easier to reason about.

02

Local-AI focused

Built to make local-AI decisions easier to reason about.

03

Interactive hero

Built to make local-AI decisions easier to reason about.

04

Runyard design system

Built to make local-AI decisions easier to reason about.

05

Two or more target models

Grounded in the actual inputs and outputs this page is designed around.

06

Strength-versus-weakness matrix

Grounded in the actual inputs and outputs this page is designed around.

07

Built for decisions, not just stats

Grounded in the actual inputs and outputs this page is designed around.

08

Standalone tool

Grounded in the actual inputs and outputs this page is designed around.

Spotlight

The differentiator behind model comparison matrix.

7B vs 14B speed

~12 tok/s (14B)~24 tok/s (7B)2× faster locally

7B vs 14B quality

Lower ceilingHigher ceilingSize wins on quality

Practical local pick

Always bigger7B in most casesSpeed wins daily use

Visual comparison

Clarity

Fit

Actionability

Reading Results

How to read the output tiers.

Comfortable

<70%

Enough breathing room for normal use.

Tight

70%-95%

Should work, but overhead matters.

Borderline

95%-110%

Likely needs one tradeoff.

Too heavy

>110%

Time to step down.

Quick Reference

Common setups at useful defaults.

Scenario	Baseline	Result	Notes
Starter setup	7B / Q4 / 8K	Light local target	Good first benchmark
Balanced setup	8B / Q4 / 16K	Everyday sweet spot	Works for many users
Heavier setup	14B / Q5 / 16K	Quality-focused target	Needs stronger hardware
Stretch setup	32B / Q4 / 16K	Ambitious local target	Useful upper bound

* These are approximations for planning, not a promise of exact runtime behavior.

Benefits

Why people use model comparison matrix.

01

Faster decisions

It helps eliminate dead-end local AI choices before you download, benchmark, or configure too much.

02

Clearer tradeoffs

The page turns a raw estimate into something you can actually act on.

03

Useful on its own

The hero provides a working tool surface while the rest of the page explains what the output means.

FAQ

Questions people ask before using model comparison matrix.

What criteria matter most when comparing local models?

Fit (does it load without OOM), speed (tok/s on your hardware), context capacity, and output quality. Fit and speed are hardware-dependent; quality is model-dependent. Fit is always the first gate.

How do I choose between a 7B and a 14B model?

If both fit comfortably, start with 14B for quality-sensitive tasks. If the 14B runs below 10 tok/s and the 7B runs at 25+ tok/s, the 7B likely wins for everyday use despite its lower theoretical ceiling.

Does model family matter more than size?

For general tasks, size usually dominates. But some families punch above their weight — Qwen 2.5 14B often outperforms older 70B models on benchmarks. Check specific use-case benchmarks, not just parameter counts.

What is the difference between instruction-tuned and base models?

Instruction-tuned models are fine-tuned to follow chat prompts and return helpful responses. Base models are raw pre-trained weights — useful for fine-tuning but not for direct chat without careful prompting.

How do I compare long-context quality across models?

Use a consistent test: give a document plus a detail question, run each model at the same context length, and compare accuracy and coherence. Practical testing beats abstract benchmark scores for your specific use case.

Where can I run a live model comparison for my hardware?

Runyard Compare lets you put two models side by side and see how they score on your GPU class. Model Radar shows all models ranked by hardware fit — both tools are better than comparing spec sheets.

RUNYARD.DEV / Tools / Model Comparison Matrix

Estimates on this page are directional and should be validated against your actual runtime and hardware.