Mistral AI Launches Forge to Train Frontier Models on Proprietary Data
· Updated
Mistral AI launched Forge, a system for enterprises to build frontier-grade AI models on proprietary data. Forge supports pre-training on internal document and codebase collections, post-training to refine model behavior, and reinforcement learning to align models with internal policies. It works across dense and mixture-of-experts (MoE) architectures, handles multimodal inputs, and is agent-first — AI agents can drive fine-tuning and hyperparameter search using plain English.
Generic public models lack the institutional context enterprises depend on. Forge closes that gap by encoding organizational terminology, compliance rules, and operational processes directly into model behavior. This makes enterprise agents more reliable: tool selections become more precise, multi-step workflows more dependable, and outputs consistent with internal governance.
Sign up to explore whether building a custom model fits your organization's needs. Enterprises can bring internal documentation, codebases, and operational records to build models that reason in domain language.
Today, we’re introducing Forge, a system for enterprises to build frontier-grade AI models grounded in their proprietary knowledge.
🌎 Forge bridges the gap between generic AI and enterprise-specific needs. Instead of relying on broad, public data, organizations can train models that understand their internal context embedded within systems, workflows, and policies, aligning AI with their unique operations.
We have already partnered with world-leading organizations, like ASML, DSO National Laboratories Singapore, Ericsson, European Space Agency, Home Team Science and Technology Agency (HTX) Singapore and Reply to train models on the proprietary data that powers their most complex systems and future-defining technologies.
