Mistral AI Releases Forge
New platform trains AI on internal docs and workflows, moving beyond generic models for complex enterprise tasks.
Mistral AI has launched Forge, a new platform designed to help enterprises build custom, frontier-grade AI models trained on their proprietary internal data. Unlike generic models trained on public information, Forge enables organizations to ground AI in their unique institutional knowledge—including engineering standards, compliance policies, codebases, and operational records. The system supports the full model lifecycle, from pre-training on large internal datasets to post-training refinement and reinforcement learning for aligning models with specific policies and objectives. This approach allows the resulting models to learn the specific vocabulary, reasoning patterns, and constraints of the enterprise environment.
Forge is already being used by major organizations like semiconductor giant ASML, telecom leader Ericsson, and the European Space Agency to train models on the data that powers their most complex systems. The platform offers strategic autonomy by letting companies retain full control over their models and the proprietary knowledge encoded within them, a critical feature for regulated industries. It supports multiple model architectures, including dense and mixture-of-experts (MoE), for flexibility in performance and cost.
The ultimate goal is to create more reliable and capable enterprise AI agents. Agents powered by these custom models can move beyond simple question-answering to precisely navigate internal tools, execute multi-step workflows, and make decisions that reflect internal business logic and policies. This shifts AI from being a generic assistant to an operational component integrated into core enterprise systems.
- Enables training of frontier AI models on proprietary enterprise data like internal docs and codebases, moving beyond generic public models.
- Supports full model lifecycle including pre-training, post-training, and reinforcement learning for policy alignment within enterprise workflows.
- Provides strategic autonomy and control, with early adopters including ASML, Ericsson, and the European Space Agency.
Why It Matters
It enables enterprises to build AI that truly understands their internal operations, making AI agents reliable components of complex business systems.