ML Model Raises $250 Million for Its Banking Agents

Model ML secures $250 million to deploy its AI agents in banks, with a valuation of $1.5 billion. The company combines specialized open-source models with a private installation. However, its performance claims remain distinct from an independent measure of errors.
Banks Keep Their Documents In-House
Model ML offers to install its platform in a private cloud environment or on the institution's own servers. Documents do not leave the banking network, while regulations prohibit exposing sensitive transactional data to shared external infrastructures. The company contractually excludes any use of customer data for training purposes.
The platform holds ISO/IEC 27001:2022 certification as well as a SOC 2 Type II audit under the AICPA framework, focusing on the continuous processing of sensitive information. However, no independent oversight currently quantifies the errors of the financial models that are automatically updated in the background.
In addition to this cross-functional assistant, institutions also have specialized offerings, including Eunice, Gradient Labs, and Rogo.
Model ML Claims Lower Costs
Model ML claims a cost limited to 2% of that of the most advanced proprietary systems, as well as a fivefold faster execution of its banking agents. These comparisons remain company statements. Extracting structured data mobilizes fewer resources than complex financial reasoning involving multiple steps.
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.
Choisissez votre rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
For ongoing processing, generalist models incur high computational expenses. The orchestration engine examines each operation and selects its processing based on accuracy, response time, and cost, without relying on a specific model. Repetitive office operations do not necessarily go through large proprietary models. An increasing fraction of processing is handled by open-source models that Model ML adapts from financial corpora, limiting calls to OpenAI, which are charged per million tokens, as well as dependence on models billed according to Anthropic's API pricing.
Capital Will Also Fund Banking Connectors
General Atlantic and QED Investors are jointly leading the funding for this London-based company founded by Chaz Englander and Arnie Englander. The funds raised since its inception now total $350 million. FT Partners led its Series A round of $75 million in November 2025.
For the third quarter of 2026, Model ML reports a twelvefold increase in its annual recurring revenue. The company also notes a twenty-sixfold increase in daily active users during this period.
Analysts can request updates to an Excel spreadsheet or the preparation of a preliminary audit report. The agents access internal databases and intervene directly in Excel or PowerPoint documents. In both London and New York, teams plan to enhance these connectors with the new capital.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.