Junie Local on NVIDIA RTX Spark: Free Local AI Coding
Blog post from JetBrains
JetBrains has expanded Junie Local, its free local AI coding agent, with day-one support for NVIDIA RTX Spark systems and stable Windows support for NVIDIA RTX GPUs with at least 24 GB of VRAM. Running entirely on-device, Junie Local can read and edit code, execute tests, and avoid token charges, quotas, and external code transmission. RTX Spark’s Blackwell GPU and Grace CPU enable local model inference speeds of up to 55 tokens per second in JetBrains’ initial testing, while adaptive context sizing, FP8 Tensor Core computation, and speculative decoding improve performance. The updated installer automatically checks compatible hardware and drivers, installs an optimized engine and model, and supports Qwen-based model options; users need Windows 11, NVIDIA driver version 581 or later, and roughly 20 GB of disk space. JetBrains plans further model, memory-efficiency, coding-performance, and RTX Spark inference improvements, and is inviting feedback from users on supported hardware.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| AI Coding Assistant | 3 | 953 | 270 | 99 | -40% |
| AI Agents | 1 | 3,062 | 780 | 195 | -45% |
| LLM | 1 | 3,134 | 613 | 172 | -48% |
| Local AI | 1 | 108 | 36 | 14 | -50% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.