Apertus in Lumo

Apertus 1.5 is one of the open-source AI models available in Lumo, Proton’s privacy-first AI assistant. Developed as an independent European alternative by the researchers at  EPFL, ETH Zurich, and the Swiss National Supercomputing Centre (CSCS), Apertus has an open architecture and is trained on publicly available data, respecting opt-out requests and filtering personal details. 

Proton is partnering with the researchers behind Apertus to help improve its model through optional, real-world feedback. If you choose to rate an Apertus response and submit feedback, your contribution is shared with Proton and the research teams behind the model.

How to use Apertus 1.5 in Lumo

1. Go to lumo.proton.me(neues Fenster) or open the Lumo mobile app.

2. Under the chat window, select the model icon at the bottom-right corner.

3. In the modal, select Apertus 1.5 as the model option.

4. Read the introduction to Apertus in the pop-up window, then select Continue.

All done. Lumo will use the Apertus 1.5 model for all responses until you change it.

How to share feedback on Apertus responses in Lumo

Sharing feedback helps Apertus’s research teams improve the model. It is optional, and all feedback is anonymized before it is shared.

1. Under an Apertus response, select the thumbs-up or thumbs-down icon to rate it as helpful or unhelpful.

2. If you selected the thumbs-down option, you’ll be asked to opt in to share your feedback with Proton and the Apertus team. Select Continue to proceed.

3. Fill out the feedback form, then select Submit.

All done. Your anonymized contribution will be sent to help the Apertus team identify areas for improvement.