Mistral Forge: Enterprise Platform for Building Custom AI Models with Reinforcement Learning
Mistral AI announced Forge at Nvidia GTC 2026 on March 17 — a platform that lets enterprises build frontier-grade AI models trained on their own proprietary data.
Forge supports the full model training lifecycle: pre-training on large internal datasets, post-training through supervised fine-tuning, DPO, and ODPO, and reinforcement learning pipelines designed to align models with internal policies and evaluation criteria. It also includes data pipeline capabilities for data acquisition, curation, and synthetic data generation.
The platform supports dense and Mixture-of-Experts architectures with multimodal input, and features an agent-first design where autonomous agents like Mistral Vibe can fine-tune models, optimize hyperparameters, schedule jobs, and generate synthetic data. Early partners include ASML, Ericsson, European Space Agency, and DSO National Laboratories Singapore.
Mistral says it is on track to surpass $1 billion in annual recurring revenue this year. Forge represents a pointed move to own the enterprise custom model market, positioning Mistral as the infrastructure layer for organizations that want AI grounded in their own knowledge rather than generic public models.
Official announcement: https://mistral.ai/news/forge
← Back to all articles
Forge supports the full model training lifecycle: pre-training on large internal datasets, post-training through supervised fine-tuning, DPO, and ODPO, and reinforcement learning pipelines designed to align models with internal policies and evaluation criteria. It also includes data pipeline capabilities for data acquisition, curation, and synthetic data generation.
The platform supports dense and Mixture-of-Experts architectures with multimodal input, and features an agent-first design where autonomous agents like Mistral Vibe can fine-tune models, optimize hyperparameters, schedule jobs, and generate synthetic data. Early partners include ASML, Ericsson, European Space Agency, and DSO National Laboratories Singapore.
Mistral says it is on track to surpass $1 billion in annual recurring revenue this year. Forge represents a pointed move to own the enterprise custom model market, positioning Mistral as the infrastructure layer for organizations that want AI grounded in their own knowledge rather than generic public models.
Official announcement: https://mistral.ai/news/forge