compar:IA

compar:IA is a chatbot arena developed within the French Ministry of Culture and Interministerial Directorate of Digital Affairs with a dual mission:

• To raise awareness about the diversity of models, cultural and linguistic biases, and the environmental impact of conversational AIs.

• To create and publish datasets of questions and user preferences useful for observing usage and improving French-language conversational AIs.

The Compar:IA tool is developed as part of the State startup compar:IA (incubated by the Atelier numérique and AllIAnce) integrated into the beta.gouv.fr program of the Interministerial Digital Directorate (DINUM), which helps public administrations build useful, simple, and user-friendly digital services.

Links:

• ✨ Try compar:IA

• 📧 Contact us

• 💻 Code repository

Datasets

The data collected through compar:IA is published in three datasets:

💬 comparIA-conversations

Dataset containing all the questions asked and answers received on the compar:IA platform.

379k+ conversations

👍 comparIA-votes

Dataset grouping user preferences expressed after a full conversation. These votes compare two models when no individual message reactions are given.

127k+ votes

🫶 comparIA-reactions

Dataset collecting user reactions to compar:IA at the message level. It reflects preferences expressed throughout conversations, message by message.

78k+ reactions