Open recommendation engine · registry-verified 100% local · no API keys

Find the Best Local AI Model for Your Device

Tell us your hardware and get the single best local LLM it can actually run: one verified pick sized to your exact memory, not a generic chatbot guess. Honest speed and quality estimates for Apple Silicon, iPhone, NVIDIA, and AMD.

Or jump to any device, try: RTX 4090, MacBook Pro M5, 16GB, iPhone 17 Pro

107
models tracked
75
run locally via Ollama
51
devices & GPUs covered
Hardware config
·
·
STEP 01

Choose your device

We match models to your exact memory and architecture.

Command line

Install the optimized LLM for your machine in one line

ModelFit's free, zero-dependency CLI detects your exact hardware, picks the single best local model it can run, and hands you the one-line Ollama command to install it. The dataset behind it is open (CC BY 4.0), and AI agents can query it live over MCP.

NPM · @wecko-ai/modelfit · MIT
$npx @wecko-ai/modelfit
Field notes

Latest from the blog

View all
Key facts

How much LLM fits your memory?

All stats

A local LLM needs roughly 0.6 GB per billion parameters at Q4. ModelFit sizes picks to ~70% of your unified memory on machines up to 32GB, scaling to ~85% at 128GB+, leaving room for the OS, context, and KV-cache. Here is what each tier comfortably runs today. The same math applies to GPU VRAM: for dedicated cards, see the per-card GPU guides.

Unified memoryRuns up toTop local pickModels that fit
8 GB~8.3B paramsLFM2.5 8B-A1B23 of 75
12 GB~12B paramsGemma 4 12B29 of 75
16 GB~14B paramsQwen3.5 9B Instruct (Q8)37 of 75
24 GB~27B paramsGemma 4 12B (Q8)45 of 75
32 GB~35B paramsQwen3.6 35B-A3B54 of 75
36 GB~35B paramsQwen3.6 35B-A3B56 of 75
48 GB~46.7B paramsQwen3.6 27B (Q8)59 of 75
64 GB~70B paramsQwen3.6 35B-A3B (Q8)64 of 75
72 GB~80B paramsQwen3.6 35B-A3B (Q8)65 of 75
96 GB~122B paramsQwen3.5 122B-A10B Instruct70 of 75
128 GB~122B paramsQwen3.5 122B-A10B Instruct71 of 75
192 GB~235B paramsQwen3 235B A22B72 of 75
256 GB~235B paramsQwen3 235B A22B72 of 75

Derived from ModelFit's own catalog of 107 models · dataset updated 2026-07-06 · full dataset (CC BY 4.0)

Benchmarks

Local vs Cloud: How Close Are We?

Open-weight models you can run at home keep narrowing the gap with the best closed APIs on coding. Here is where the line sits today on SWE-Bench Verified.

Full benchmark
Claude Fable 5
95.0%
Claude Opus 4.8
88.6%
GPT-5.5
82.6%
Qwen3.6-27B
77.2%

SWE-Bench Verified · third-party scores, each raw-confirmed against the model's primary source. Qwen3.6-27B runs locally on a 24GB Mac.

Explore

Browse the full index

All devices
FAQ

Local LLM questions, answered

How much memory do I need to run a local LLM?

A local model needs roughly 0.6 GB of memory per billion parameters at Q4 quantization. ModelFit sizes its picks to about 70% of your device's unified memory (scaling to ~85% on 128GB+ machines), so an 8GB machine comfortably runs models up to ~8.3B and a 16GB machine up to ~14B parameters.

What is the best local LLM for a 16GB Mac?

On a 16GB Mac, ModelFit's top pick is Qwen3.5 9B Instruct (Q8): it loads in about 10.7GB and is the highest-quality local model that fits comfortably. 37 of ModelFit's 75 local models fit a 16GB device.

Can I run these models completely offline?

Yes. Every model ModelFit recommends runs fully offline through Ollama on your own hardware: no API keys, no subscription, and no data leaves your machine.

How many AI models does ModelFit cover?

ModelFit tracks 107 AI models across 21 families; 75 run locally via Ollama on Apple Silicon, NVIDIA, or AMD hardware.

How accurate are the speed estimates?

Tokens-per-second figures are ModelFit estimates derived from your chip or GPU's memory bandwidth and the model's size, not measured benchmarks. Real-world speed varies with quantization, context length, and thermal state.