Apertus
Ranked 26 of 39 open-weight language models by recommenders
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.
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.
29 people gave an opinion
- recommend it
- 17
- criticize it
- 12
- approval share
- 59%
Recommended in 7 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: recommended
- February 2026: not recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: not recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it doesn’t carry the badge
- 17 people recommend itNeeds 10 or more, met.
- 1.4 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 7 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Fully open, not just open weights
“Fully open model: open weights + open data + full training details including all data and training recipes”
Comment from 21 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI” “Apertus and Soofi, both open source (not only open weight) models”
Comment from 23 Jul 2026 in “Startup founders urge Trump not to shut off Chinese open weight AI” “Great to see more fully open LLMs.”
Comment from 21 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI” Lower copyright and data risk
“It doesn't produce state-of-the-art (SOTA) results, but for many organizations, it greatly reduces risk of copyright infringement”
Comment from 11 Mar 2026 in “Open Weights Isn't Open Training” “Apterus even respects opt-out notices in their collection of the dataset they distribute”
Comment from 8 Jul 2026 in “Chatto is now Open Source” “they built it properly regarding copyright and data transparency”
Comment from 17 Jun 2026 in “GLM-5.2 is the new leading open weights model on Artificial Analysis” Valued by the research community
“By far the most impactful product of the Apretus project are the people.”
Comment from 21 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI” “they are very important for the community”
Comment from 22 May 2026 in “The current AI pricing was always going to go away” “Very cool that projects like this exist.”
Comment from 20 Apr 2026 in “Swiss AI Initiative” Competitive for its model size
“Soofi S and Apertus 1.5 outperform or are at least comparable to Ai2 depending on benchmarks.”
Comment from 8 Aug 2026 in “U.S. Department of Energy Launches the Genesis Open Models Initiative” “Apertus is very good for the model size”
Comment from 22 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI”
What people warn about
Each mark is one comment
Weak performance against rival models
“their models trail not only US models, but Chinese models, including smaller ones, by a significant amount”
Comment from 22 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI” “The LLM they were involved in last year (Apertus) still was a letdown.”
Comment from 18 Sep 2026 in “Qwen 3.8 Omni Flash” “I have not been impressed by the earlier Apertus 8B model”
Comment from 22 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI” - Comment from 22 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI”
- Comment from 22 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI”
Unreliable on simple questions
“it's quite unreliable when it comes to simple questions like "how to say X in language Y"”
Comment from 21 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI” “based on my testing, that's not true either. So in my view this is completely useless.”
Comment from 21 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI” Unlicensed web data in training
“It uses fineweb, which is derived from Common Crawl, which is an unlicensed scrape of web pages.”
Comment from 21 Jun 2026 in “Apertus – Open Foundation Model for Sovereign AI” “They didn't release any new data.”
Comment from 21 Nov 2025 in “Olmo 3: Charting a path through the model flow to lead open-source AI”
Other open-weight language models people recommend
The four most recommended other open-weight language models, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Qwen 3.8 27B | 1,173 | 426 | 73% | 13 | 2 mo32 owners | Approved |
| Z.ai GLM | 1,043 | 384 | 73% | 13 | 2 mo52 owners | Approved |
| Kimi | 931 | 359 | 72% | 13 | 1 mo26 owners | Approved |
| Gemma 4 | 713 | 249 | 74% | 13 | 1 mo14 owners | Approved |
| ApertusThis page | 17 | 12 | 59% | 7 | — | — |
What people said
29 of the 30 comments that mention it, newest first. Each links to the original comment.
“The LLM they were involved in last year (Apertus) still was a letdown.”
“Very useful if you are doing LLM research.”
“Soofi S and Apertus 1.5 outperform or are at least comparable to Ai2 depending on benchmarks.”
“HF needs auth. No thanks.”
“Apertus and Soofi, both open source (not only open weight) models”
“Apterus even respects opt-out notices in their collection of the dataset they distribute”
“I have not been impressed by the earlier Apertus 8B model”
“Apertus is very good for the model size”
“Apertus V1 performance were sub-par.”
“so it's not quite there yet.”
“their models trail not only US models, but Chinese models, including smaller ones, by a significant amount”
“direct threat to commercial AI labs and their business model”
“These models don't seem very competitive”
“By far the most impactful product of the Apretus project are the people.”
“Is a recipe useful if no one likes it?”
“Great to see more fully open LLMs.”
“the impression I get from Apertus is that it moves at the speed of a committee”
“It uses fineweb, which is derived from Common Crawl, which is an unlicensed scrape of web pages.”
“it's quite unreliable when it comes to simple questions like "how to say X in language Y"”
“It is not ready for agentic use, but as a generic driving model for RAG use cases, it is pretty competent.”
“Looks like their instruct models are Llama3.1 fine tune from last year.”
“Fully open model: open weights + open data + full training details including all data and training recipes”
“based on my testing, that's not true either. So in my view this is completely useless.”
“they built it properly regarding copyright and data transparency”
“they are very important for the community”
“Very cool that projects like this exist.”
“It doesn't produce state-of-the-art (SOTA) results, but for many organizations, it greatly reduces risk of copyright infringement”
“if you want a truly open model you can always try Apertus”
“They didn't release any new data.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.