Mistral Launches Large 4 AI Model, Tops GPT-6 Astra on One Finance Test

The claim rests on a single finance benchmark, which is a narrow basis for a competitive verdict — model families routinely trade wins across task-specific evals, and Mistral's own results show it trailing elsewhere. The bigger signal is positioning: a Paris lab framing itself against OpenAI and Anthropic on finance use cases speaks to where enterprise demand and RWA-adjacent analytics are concentrating. With no release date, pricing or parameter count disclosed, whether Large 4 becomes an accessible option rather than a leaderboard entry is the open question.