# Unsloth GGUF quants

Hacker Approved page: https://hackerapproved.com/p/unsloth-quants/

33 people recommended Unsloth GGUF quants in Hacker News comments, 2 had mixed feelings and 10 criticized it, November 2025 to October 2026. It was recommended in 10 of 12 months, in 28 threads. It meets the Hacker Approved badge rule.

**Verdict:** Widely recommended GGUF quantizations of popular open models, valued for quick releases and low memory use, though some early releases had bugs and errors.

Summary written by Claude from 48 comments.

## Why people recommend it

- Reliable, go-to quants (8 comments in the sample): "The unsloth quants seem to have been fixed, so they are probably the go-to again" ([source](https://news.ycombinator.com/item?id=47200766)) "I have had great luck with unsloth quants so far." ([source](https://news.ycombinator.com/item?id=49281117)) "They are the most reliable in my experience" ([source](https://news.ycombinator.com/item?id=49279091))
- Runs on low-memory hardware (5 comments in the sample): "something like Unsloth's UD q8 XL or q6 XL quants free up a lot of memory and bandwidth" ([source](https://news.ycombinator.com/item?id=47821402)) "amazed by the Unsloth team by how much they can reduce the requirements" ([source](https://news.ycombinator.com/item?id=50293101)) "the unsloth versions can run on a potato and are actually useful" ([source](https://news.ycombinator.com/item?id=48214796))
- Better quality than self-made quants (3 comments in the sample): "They are higher quality than the quants you can make yourself due to their imatrix datasets" ([source](https://news.ycombinator.com/item?id=47822260)) "Unsloth's quants are better than the original Google's QAT as posted in the article." ([source](https://news.ycombinator.com/item?id=48415688)) "we're now the best in terms of KLD and disk space" ([source](https://news.ycombinator.com/item?id=47795073))
- Tested, quickly published quants (3 comments in the sample): "they make it first, then they run a lot of tests, KD, perplexity, etc, they publish it" ([source](https://news.ycombinator.com/item?id=49277872)) "Unsloth are doing a fantastic job of publishing quants of the major models quickly" ([source](https://news.ycombinator.com/item?id=48139774)) "unsloth produces nice quants for it and even provides guidelines" ([source](https://news.ycombinator.com/item?id=47541277))

## What people warn about

- Errors and bugs in releases (5 comments in the sample): "I had several issues with unsloth gguf, even for models released a few months back like gemma 4" ([source](https://news.ycombinator.com/item?id=49277227)) "Sometimes unsloth has broken ones for a particular model, sometimes no quants at all" ([source](https://news.ycombinator.com/item?id=47272866)) "Running unsloth versions with llama.cpp, constant issues." ([source](https://news.ycombinator.com/item?id=48730837))
- Quality dips at lower bits (3 comments in the sample): "They seem to make less errors at lower "average" quantizations than the unsloth 4 bit quants." ([source](https://news.ycombinator.com/item?id=46684879)) "4 bit unsloth quants are good if you never ask for more than 20k context" ([source](https://news.ycombinator.com/item?id=48557579)) "They definitely feel a bit dazed." ([source](https://news.ycombinator.com/item?id=47192540))

## What people recommend instead

- [Unsloth Dynamic GGUFs](https://hackerapproved.com/p/unsloth-dynamic-2-0-ggufs/): 13 recommend, 2 mixed, 7 criticize

## The badge

A product earns the Hacker Approved badge when at least 10 people recommend it, recommenders outnumber critics at least 2 to 1 (mixed feelings count toward neither side), it was recommended in at least 4 of the 12 months and in at least 5 different threads, and no single thread supplies more than half of the people who recommend it. Health and money products are never badged.

## Cite as

Hacker Approved, "Unsloth GGUF quants", https://hackerapproved.com/p/unsloth-quants/ (33 recommend, 2 mixed, 10 criticize; Hacker News comments, November 2025 to October 2026).

Counts updated 2 Nov 2026, with data through October 31, 2026.
