# Apertus

17 Hacker News commenters recommended Apertus and 12 criticized it, November 2025 to November 2026. It was recommended in 7 of 13 months.

**Verdict:** Free fully open model with open weights, data and training details, valued for copyright safety and research. Critics say it trails US and Chinese models.

## Why people recommend it

- Fully open, not just open weights (3 comments): "Fully open model: open weights + open data + full training details including all data and training recipes" ([source](https://news.ycombinator.com/item?id=48622955)) "Apertus and Soofi, both open source (not only open weight) models" ([source](https://news.ycombinator.com/item?id=49027163)) "Great to see more fully open LLMs." ([source](https://news.ycombinator.com/item?id=48623213))
- Lower copyright and data risk (3 comments): "It doesn't produce state-of-the-art (SOTA) results, but for many organizations, it greatly reduces risk of copyright infringement" ([source](https://news.ycombinator.com/item?id=47337014)) "Apterus even respects opt-out notices in their collection of the dataset they distribute" ([source](https://news.ycombinator.com/item?id=48838940)) "they built it properly regarding copyright and data transparency" ([source](https://news.ycombinator.com/item?id=48570455))
- Valued by the research community (3 comments): "By far the most impactful product of the Apretus project are the people." ([source](https://news.ycombinator.com/item?id=48623524)) "they are very important for the community" ([source](https://news.ycombinator.com/item?id=48240084)) "Very cool that projects like this exist." ([source](https://news.ycombinator.com/item?id=47829761))
- Competitive for its model size (2 comments): "Soofi S and Apertus 1.5 outperform or are at least comparable to Ai2 depending on benchmarks." ([source](https://news.ycombinator.com/item?id=49223622)) "Apertus is very good for the model size" ([source](https://news.ycombinator.com/item?id=48629365))

## What people warn about

- Weak performance against rival models (5 comments): "their models trail not only US models, but Chinese models, including smaller ones, by a significant amount" ([source](https://news.ycombinator.com/item?id=48626836)) "The LLM they were involved in last year (Apertus) still was a letdown." ([source](https://news.ycombinator.com/item?id=49750742)) "I have not been impressed by the earlier Apertus 8B model" ([source](https://news.ycombinator.com/item?id=48633307))
- Unreliable on simple questions (2 comments): "it's quite unreliable when it comes to simple questions like "how to say X in language Y"" ([source](https://news.ycombinator.com/item?id=48623188)) "based on my testing, that's not true either. So in my view this is completely useless." ([source](https://news.ycombinator.com/item?id=48622936))
- Unlicensed web data in training (2 comments): "It uses fineweb, which is derived from Common Crawl, which is an unlicensed scrape of web pages." ([source](https://news.ycombinator.com/item?id=48623199)) "They didn't release any new data." ([source](https://news.ycombinator.com/item?id=46002918))

## Alternatives hackers recommend

- [Qwen 3.8 27B](https://hackerapproved.com/p/qwen-3-8-27b.md): 1173 recommend
- [Z.ai GLM](https://hackerapproved.com/p/glm-5-2.md): 1043 recommend
- [Kimi](https://hackerapproved.com/p/kimi-k3.md): 931 recommend
- [Gemma 4](https://hackerapproved.com/p/gemma-4.md): 713 recommend
- [Qwen](https://hackerapproved.com/p/qwen.md): 474 recommend

## Cite as

Hacker Approved, "Apertus", https://hackerapproved.com/p/apertus/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).