Twoody
Local LLM calculator
A local LLM on your Intel Mac
Twoody runs on Intel Macs with macOS 13 or later. The model runs on the processor: small models work, slowly.
| Your computer | Memory | Recommended — runs comfortably | Writes | One click in Twoody |
|---|---|---|---|---|
Intel MacBook Air, MacBook Pro 13″ (2018–2020) |
8 GB | Qwen3 0.6B | 22–65 tokens/s | Qwen3 4B |
| 16 GB | Llama 3.1 8B | 3.4–6.1 tokens/s | Qwen3 8B | |
| 32 GB | GLM-4.7-Flash | 7.9–14 tokens/s | Qwen3 14B | |
Intel MacBook Pro 15″, 16″ (2018–2019) |
16 GB | Llama 3.1 8B | 3–5.4 tokens/s | Qwen3 8B |
| 32 GB | GLM-4.7-Flash | 7.1–13 tokens/s | Qwen3 14B | |
| 64 GB | Qwen3-Next 80B-A3B | 6–11 tokens/s | Qwen3 14B | |
Intel iMac, Mac mini (2017–2020) |
8 GB | Qwen3 0.6B | 21–62 tokens/s | Qwen3 4B |
| 16 GB | Llama 3.1 8B | 3.2–5.7 tokens/s | Qwen3 8B | |
| 32 GB | GLM-4.7-Flash | 7.5–13 tokens/s | Qwen3 14B | |
| 64 GB | Qwen3-Next 80B-A3B | 6.4–11 tokens/s | Qwen3 14B | |
iMac Pro, Mac Pro (2017–2019) |
32 GB | GLM-4.7-Flash | 17–30 tokens/s | Qwen3 14B |
| 64 GB | Qwen3-Next 80B-A3B | 14–25 tokens/s | Qwen3 14B | |
| 128 GB | Qwen3-Next 80B-A3B | 14–25 tokens/s | Qwen3 14B |
Estimates, not measurements, computed from public llama.cpp benchmarks. Figures reviewed on September 26, 2026.
On an Intel Mac, llama.cpp uses the processor, not the graphics: a model of about 4 billion parameters writes a few words per second, enough for short answers.
Twoody installs the Qwen3 that suits the memory; on these Macs, the 4B is the one that answers fastest.
LM Studio needs an Apple Silicon Mac and macOS 14, and Ollama macOS 14: on an Intel Mac or with macOS 13, Twoody installs and runs a model in one click.
Questions about running an LLM locally
Does a local LLM work on an Intel Mac?
Yes, on the processor: a model of 4 billion parameters writes a few tokens per second, and reading a long document takes minutes.
Which local AI app works on macOS 13?
Twoody runs on macOS 13 (Ventura) and later, on Apple Silicon and Intel; it installs a model in one click.
Which model for an Intel Mac?
The smallest ones that are still useful: Qwen3 4B or Qwen3.5 4B. Larger models fit in memory but write too slowly for comfort.
The best local LLM, by computer
Try local AI on your Mac, for free.
Version 0.11.5 · macOS 13 or later · no account, no subscription. Download it, install a model in one click, and ask your first question — even offline.