Releases·Tool Release
Solo developer releases Jenny, a free MIT-licensed Electron desktop app for running local LLMs with tool calling
A developer released Jenny, a free MIT-licensed Electron desktop app built over more than a year of nights and weekends. It runs local LLMs with tool calling, rollback, and an IDE. The author built it to enable private, local hosting ahead of any end to subsidized frontier LLM pricing.
214 engagement1 sourceMon, Sep 7, 2026, 04:27 PM
AI Overview
- Jenny is a free, MIT-licensed Electron desktop app built solo over 1.5 years for running local LLMs privately with tool calling, rollback, and a full IDE
- The app supports llama.cpp, vLLM, or any OpenAI-compatible endpoint, plus GGUF via a managed llama-server, and recommends ornith1.5:9b for tool calling and speed
- Safety features include approval for destructive shell commands, checkpointed file edits for rollback, and no network calls except to the local model runtime
- Windows is the only tested build (tested on an RTX 5070 Ti); macOS is untested, no Linux build, and the installer .exe is unsigned so SmartScreen warns users
- The developer cites subsidized frontier LLM pricing from venture capital as motivation and points to existing alternatives like LM Studio, Open WebUI, and Unsloth Desktop
Related
Products·launched
Local LLM workspace with hardware awareness, agents, skills, MCP
local-llmup is a desktop app for macOS, Windows, and Linux that runs local LLMs with hardware awareness, supporting agents, skills, and MCP, all fully local and free. It helps users avoid vendor lock-in and choose the right model for their hardware.
local-llmup is a desktop app for macOS, Windows, and Linux that runs local LLMs with hardware awareness, supporting agents, skills, and MCP, all fully local and free. It helps users avoid vendor lock-in and choose the right model for their hardware.
0 engagement1 sourceredditThu, Aug 27, 2026, 02:07 PM
Products·launched
Ember - Local AI: run LLMs on your phone or connect to local servers
Ember - Local AI is a free Android app that lets users run local AI models on their phones without an internet connection, or connect to locally running servers. It addresses privacy concerns and offline usability for LLM users, with iOS support coming soon.
Ember - Local AI is a free Android app that lets users run local AI models on their phones without an internet connection, or connect to locally running servers. It addresses privacy concerns and offline usability for LLM users, with iOS support coming soon.
2 engagement1 sourceredditFri, Aug 14, 2026, 11:16 AM
Products·building
Vyact: open-source desktop workspace for local LLMs, documents, browser context, and coding
Vyact is a free, open-source desktop app that combines local model management, document RAG, browser context, and coding workflows in one place. It handles downloading and running local models (using MLX on Apple Silicon) so users can work with their own files alongside a local LLM. It targets people who want everyday AI tasks without juggling separate tools.
Vyact is a free, open-source desktop app that combines local model management, document RAG, browser context, and coding workflows in one place. It handles downloading and running local models (using MLX on Apple Silicon) so users can work with their own files alongside a local LLM. It targets people who want everyday AI tasks without juggling separate tools.
3 engagement1 sourceredditFri, Sep 4, 2026, 03:12 PM
Products·launched
HilbertRaum: private local AI chat and document analysis
HilbertRaum is an open-source desktop app that runs LLMs entirely on your own computer, enabling private AI chat and document analysis. It is designed for non-technical users, with features like source-citing answers and encrypted workspace, ensuring no data leaves the machine.
HilbertRaum is an open-source desktop app that runs LLMs entirely on your own computer, enabling private AI chat and document analysis. It is designed for non-technical users, with features like source-citing answers and encrypted workspace, ensuring no data leaves the machine.
0 engagement1 sourceredditTue, Jul 21, 2026, 04:25 PM
Products·launched
Openlumara: Lightweight modular AI agent webUI for local models
Openlumara is a fully open-source (GPL3) AI agent framework and webUI designed for local models, with no subscriptions or corporate backing. It offers a lightweight, modular alternative to OpenClaw and Hermes, and its webUI can be used standalone for chatting with local models. The recent rewrite removed AI-generated code and switched to Alpine.js for a cleaner, more maintainable codebase.
Openlumara is a fully open-source (GPL3) AI agent framework and webUI designed for local models, with no subscriptions or corporate backing. It offers a lightweight, modular alternative to OpenClaw and Hermes, and its webUI can be used standalone for chatting with local models. The recent rewrite removed AI-generated code and switched to Alpine.js for a cleaner, more maintainable codebase.
151 engagement1 sourceredditThu, Aug 6, 2026, 07:24 AM