JetBrains' Junie Local Adds Day 0 Support for NVIDIA RTX Spark
JetBrains has updated its local AI coding tool, Junie Local, to feature Day 0 support for the NVIDIA RTX Spark platform while upgrading its Windows support to stable. The update delivers generation speeds of up to 55 tokens per second while keeping all code processing completely offline. This integration provides developers with high-performance, privacy-focused local AI assistance without cloud subscription costs, token fees, or usage quotas. It significantly lowers the barrier for running advanced coding models locally on capable desktop and laptop hardware. Junie Local requires hardware with at least 24GB of VRAM and an Ampere or newer RTX GPU, or an RTX Spark device, along with Windows 11 and an updated NVIDIA driver. Users can choose between models such as Qwen3.6-27B or a proprietary JetBrains fusion model, with automatic context scaling based on available GPU memory.
## BACKGROUND
Nvidia RTX Spark is an Arm-based system-on-chip and computing platform developed by Nvidia and Microsoft for Windows laptops and compact desktop computers, combining a Grace CPU with a Blackwell RTX GPU and unified memory. Junie Local is a feature within JetBrains' AI coding tools that allows developers to run large language models locally on their own machines for code completion and editing.