Case study ·
Banking group BPCE closely benchmarks its AI models
Original title, in French: “Le groupe bancaire BPCE compare de près ses modèles d’IA”
- Publication
- La Revue du Digital (France)
- Author
- Alain Clapaud
- Published on
- 6 October 2026 · read in the page metadata · updated on 6 October 2026
Summary
Banking group BPCE is seeking to choose the right large language model (LLM) at the right cost for its various business uses, amid fierce competition between AI models. Given the limits of public benchmarks, which it considers saturated and not representative of the constraints specific to the banking industry, BPCE is setting up its own evaluation framework. It measures the cost, latency, robustness and business relevance of the models, in order to avoid silent regressions when the models in use are updated.
Summary based on the full text of the article.
Read the article on La Revue du Digital →Classification
- Country
- France
- Company size
- Large companies
- Industries
- Banking, insurance, asset management, finance departments
- Companies
- BPCE (deploys AI)
- Keywords
- language models, banking, LLM evaluation, benchmark, cost of AI, generative AI
Entry written by Marsanne Studio from the cited article, with the help of an AI model (claude-sonnet-5). The rights to the article and its illustrations belong to La Revue du Digital and its author. Entry translated from French by Claude Sonnet 5.5.