Grux AI Download

How fast is a local LLM on a Mac?

On a 16 GB Mac mini with an Apple M4, the Grux Local 2B writes about 48.0 tokens a second, the 4B about 24.6 and the 9B about 18.3, measured on 17 November 2026 through Ollama 0.40.8. A token is roughly three quarters of a word, so the 4B writes about 18.5 words a second, where a person reads about 4. With the model already loaded, the first word starts in well under a second at every size, and a 200 token reply, about 150 words, finishes in 4.2 seconds on the 2B, 8.3 on the 4B and 11.1 on the 9B. Speed falls as the model grows: the 2B is 2.6 times faster than the 9B. These are one Mac, one runner and short replies. A longer prompt, a longer context or a busier Mac reads slower, and a 24 GB Mac, the size the 9B is made for, was not measured.

Download Grux AI 3.3.1, free, for macOS 14 or later on Apple silicon.

Grux AI 3.3.1 · last checked 2026-11-13 · generated from the shipping release

Speed by size

Each row is the median of 10 timed replies from a model already in memory, on a 16 GB Mac mini with an Apple M4, macOS 26.3.1. Speed is the tokens Ollama generated divided by the time Ollama spent generating them, so it leaves out loading and reading the prompt. Across the 10 replies the speed stayed within 47.7 to 48.4 on the 2B, 22.6 to 25.8 on the 4B and 17.9 to 18.4 on the 9B, and the median time to the first word was 0.05, 0.09 and 0.12 seconds.

SizeTokens a second200 tokens
2B, for 8 GB48.04.2 s
4B, for 16 GB24.68.3 s
9B, for 24 GB18.311.1 s

Against reading speed

Tokens are not words, so this table converts at three quarters of a word a token, a rule of thumb for English, and compares with a reader at 240 words a minute, which is 4 words a second.

SizeWritesAgainst a reader at 240 words a minute
2B36.0 words a second9.0 times
4B18.5 words a second4.6 times
9B13.7 words a second3.4 times

What a wait looks like

How it was measured

What this does not say

Where to check this

ClaimSource
Speed is the tokens generated over the time spent generating themOllama API reference, the eval_count and eval_duration fields
The three tags and what each is made forGrux Local
Which Ollama models fit which memoryOllama models and which LLM can my Mac run

Questions

How many tokens a second does a local LLM make on a Mac?
On a 16 GB Mac mini with an Apple M4, about 48.0 for the Grux Local 2B, 24.6 for the 4B and 18.3 for the 9B, measured on 17 November 2026. Another Mac will differ.
Is a local LLM fast enough to use?
On the Mac mini the 4B writes about 18.5 words a second, against about 4 for a reader. The table above has the other two sizes.
Why is the first reply slower?
The model has to be read from disk into memory once. The 4B took 4.0 seconds to load on the Mac mini, and the replies after it did not pay that.
Does a bigger model run slower?
Yes. On the Mac mini the 2B was 2.6 times faster than the 9B.
Download Grux AI 3.3.1 Read the source

Free, MIT licensed. macOS 14 or later, Apple silicon. 29.1 MB, signed and notarized by Apple. No account, no server, no subscription.