Microsoft unveils MAI-Thinking-1, its advanced reasoning AI model
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisis ton rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
Microsoft took advantage of the Build 2026 event to announce a series of new artificial intelligence models, including the flagship model MAI-Thinking-1. This advanced reasoning model represents a significant milestone for the company, which had previously relied heavily on OpenAI technologies. This announcement comes after a renegotiation of their agreement, aimed at loosening the ties between the two tech giants.
MAI-Thinking-1 is described by Microsoft as a medium-sized model capable of competing with leading models in the software engineering field. The company emphasizes that this model was trained from the ground up on clean data, without resorting to distillation from third-party models, showcasing its commitment to developing robust internal solutions.
Among the other models unveiled, MAI-Image 2.5 stands out for its text-to-image conversion and image editing capabilities. MAI-Transcribe-1.5, on the other hand, promises transcription speeds five times faster than its competitors. Additionally, MAI-Voice-2, with its upcoming flash version, introduces 15 new languages and additional voice options.
Finally, the coding model MAI-Code-1-Flash, integrated with GitHub Copilot and Visual Studio Code, aims to be particularly efficient in inference, thereby strengthening Microsoft's development ecosystem.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.