This is the right terminology. Model weights are literally compiled binary data; they are the output of an algorithm run on a bunch of source data. That training dataset is the "source" of the model. Training data (or the scripts used to generate it) is human-readable and modifiable, like source code. Binary weights are not.

Just to note though, source copyright extends to its compiled form. There is probably an analogue there for model weights.

Tell me about the companies that own the copyrights to their training data.

Binary weights can still be "edited" with additional training.

Binary executables can also be edited after compilation.

Binary images can also be edited after they are created.

I propose that from now on we call freewares "open binaries".

Does an “open source” model the way you describe it exist or is it a mythical creature?

Unicorns also don't exist, but we don't change the definition to include horses.

The first level argument is that old horse, burden of proof.

An open source model does exist now [1] and is multilingual. Previous discussion [2].

[1] https://ethz.ch/en/news-and-events/eth-news/news/2025/07/a-l...

[2] https://news.ycombinator.com/item?id=44535637

It does, but does it matter? Even if every software released in 2025 was proprietary, doesn't make their published binaries "open source" because no other software could be classified as "open source".

We name things based on what they are, not based on the lack of other things.

There aren’t many but they do exist. OLMo for example.

Olmo by AllenAI and Pythia by EleutherAI.

Apertus by EPFL and ETH Zürich.

I believe it does. OLMO.

I'm not a lawyer, but I believe that the weights aren't subject to copyright. So, you can use them outside of Meta's license agreement provided you get them from somewhere else.