This content originally appeared on DEV Community and was authored by Siri Varma Vegiraju
AI/ML Enhancements
Docker Model Runner improvements include an inspector for AI inference requests and responses, allowing developers to debug model behavior by examining HTTP payloads, prompts, and outputs. The update also adds real-time resource checks to prevent system freezes when running multiple models concurrently, with warnings for GPU and memory constraints.
Expanded MCP (Model Context Protocol) support now includes Goose and Gemini CLI as clients, providing one-click access to over 140 MCP servers like GitHub, Postgres, and Neo4j through the Docker MCP Catalog.
Developer Experience
New Kubernetes CLI command (docker desktop kubernetes) allows managing Kubernetes clusters directly from the Docker Desktop CLI without switching between tools or UI screens.
Improved Settings search helps users find configurations faster without navigating through multiple menus.
Platform Performance
Apple Virtualization is now the default backend on macOS (QEMU support fully removed), delivering better performance with faster cold starts and more efficient memory management.
WSL2 improvements on Windows include reduced memory consumption, smarter CPU throttling for idle containers, and better stability for graphics-heavy workloads.
The release focuses on enhanced reliability, better AI development tools, and streamlined workflows to help developers build and test applications more efficiently, particularly those working with AI/ML models.
This content originally appeared on DEV Community and was authored by Siri Varma Vegiraju