CoreWeave Launched Forge AI Development Platform

The GPU infrastructure firm introduced a software layer to manage the full lifecycle of model and agent development.

Updated on Sept. 30, 2026 in Software

Isometric editorial illustration showing stacked steel server modules and conduit cables, representing complex AI infrastructure management. — CoreWeave introduced its Forge software platform on September 30, 2026, creating a unified management layer for the full AI development lifecycle. AI Illustration. Upload story photo >

Live Poll

Do you believe new AI development platforms make your work easier or more complex?

CoreWeave officially launched its Forge software platform on September 30, 2026, marking the company’s expansion beyond raw infrastructure services. The platform is now available to users and offers an integrated environment for managing the AI development loop.

Why it matters

By providing a software layer that connects model serving, monitoring, and evaluation, CoreWeave aims to unify fragmented workflows for AI developers. This shift moves the company from pure hardware provision into the software orchestration market.

Forge organizes AI development into five distinct stages: run, observe, curate, improve, and evaluate. The platform provides integrated tools including Agent Lens, Registry, Sandboxes, Notebooks, Training, Inference, and ARIA.

The players

CoreWeave

An infrastructure firm specializing in GPU-accelerated cloud computing for large-scale AI and high-performance computing workloads.

MasterClass

An online education platform that is an early user of the Forge software ecosystem.

Canva

A visual communication and design software company utilizing the new Forge platform.

The details

Forge functions as a central management layer that stitches together heterogeneous cloud environments into a singular workflow for AI development. It integrates specialized tools like Agent Lens—a module for monitoring agent performance—and Registry to handle the storage and versioning of machine learning artifacts. By consolidating these disparate tasks, Forge seeks to bridge the gap between underlying GPU infrastructure and high-level model lifecycle management.

June 30, 2026: CoreWeave closed its financial quarter with $2.575 billion in revenue.
September 30, 2026: The company launched the Forge platform at the Fully Connected 2026 conference in San Francisco.

The Tech Race

CoreWeave is evolving beyond its roots as a GPU-heavy cloud provider to compete with integrated AI development ecosystems. This move mirrors the competitive strategies of established players like NVIDIA DGX Cloud that seek to lock in developers by owning the entire software stack.

The Forge platform is available immediately, offering a free tier alongside a Pro version starting at $60 per month. Developers can access the integrated suite of tools to manage training and inference tasks directly within the CoreWeave environment.

The takeaway

CoreWeave's move signals that infrastructure providers are increasingly prioritizing software accessibility to retain AI developers. Watch for future integration updates to the ARIA and Agent Lens toolsets that may further differentiate the platform from legacy cloud competitors.