Maybe it would look differently if Asian nations would contribute to the funding:

https://cordis.europa.eu/project/id/825303

Nor does it include African or native South American languages, right?

Even if this had not been sponsored by the EU, it wouldn't be surprising. This kind of product relies on decades of development. The smaller and poorer a language, the less research will have been done. Chinese and Japanese often have material available, but NLP research in other countries is still behind.

Something is better than nothing. https://blog.jaibot.com/the-copenhagen-interpretation-of-eth...

It's funded by the EU and it doesn't even cover all EU languages. This is just a small scale pilot project, it's unreasonable to expect them to cover 3000 languages.

Agreed. All of the languages I use semi-regularly are in SEA. All of the open source translation options tend to neglect Asian languages.

Google Translate seems to require an account or something on my microG'd phone and refused to work after an update, which meant I actually switched to Yandex--purely because I have no other services with them so translations are somewhat siloed away from other applications and services.