How Moneo works

Moneo is Expensum's AI assistant 🤖. Here you'll find which models we use, what we use them for, and their limitations, in line with the transparency obligations of the AI Act (EU Regulation 2024/1689, Art. 50).

You are talking to an artificial intelligence

Every response from Moneo — on the Telegram bot, in the web chat, and in monthly and weekly reports — is generated automatically by a language model (LLM), not by a person. We always disclose this at the first interaction and in every report.

The AI models we use
All users

Mistral Small 3.1

Reading and categorizing expenses you write or photograph (receipts).

All users

Mistral Large 3

Conversation with Moneo in chat: answers your questions about your finances.

PRO plan only

Gemini 2.0 (Vertex AI)

Transcribing voice messages into text, before expense analysis.

All models used for text/photo analysis are provided by Mistral AI, a European company compliant with GDPR. Gemini 2.0 (Vertex AI) is used only for voice transcription, exclusively for PRO users.

Known limitations

Moneo can make mistakes: wrong categories, amounts misread from a hard-to-read receipt, or inaccurate voice transcriptions. Always verify important financial data before making decisions based on reports or Moneo's answers. You can correct any expense or income entry at any time.

Personal data and privacy

To find out how we handle your personal data and how to exercise your GDPR rights, see the privacy and data page.