Runway is introducing Model Fine-tuning, a new self-serve feature that will allow users to customize its generative models using their own datasets. The feature aims to solve the problem of general-purpose models failing to meet specific customer needs.
Key Details:
Custom training - Users will be able to fine-tune models for their specific use cases, aesthetics, or motion dynamics, moving beyond the "opinionated" nature of current general models.
Enterprise focus - Runway is highlighting applications for industries including robotics, life sciences, education, architecture, and design, suggesting a strong push toward specialized commercial use.
Pilot program access - The feature is available now for select pilot partners. A general release is "coming soon," and interested users can submit for early access.
Accessible post-training - Runway states its goal is to "make post-training accessible to everyone," allowing users to customize models with "minimal compute and data requirements."
This move positions Runway to support more specialized enterprise workflows, addressing what it identifies as a gap where general models "can fall short when deployed for real customer problems."
NVIDIA used its SIGGRAPH announcements to line up a wave of creative software behind Model Context Protocol (MCP), the open standard that lets an AI agent read an application
NVIDIA's Synthetic Video Detector Puts a Deepfake Score Inside Newsroom Workflows July 20, 2026 NVIDIA introduced the Synthetic Video Detector NIM microservice at SIGGRAPH, adding an AI-assisted signal that flags whether a video clip contains synthetic content. It is part of
Alibaba's Qwen team has opened preview access to Qwen3.8-Max , a 2.4-trillion-parameter model the company describes as its most capable system yet, with open weig
