#3838th of 73 in Desktop AI apps

locally-uncensored

Desktop app for local chat, image and video generation, and a coding agent

Releases and problems feed (RSS)

Stars
4.4k
License
AGPL-3.0
Last commit
Nov 2026
Last release
Nov 2026
Language
TypeScript
Stars, last 29 weeks+4,099
28 Apr 2026: 318 stars5 May 2026: 368 stars12 May 2026: 454 stars19 May 2026: 520 stars26 May 2026: 548 stars2 Jun 2026: 586 stars9 Jun 2026: 617 stars16 Jun 2026: 723 stars23 Jun 2026: 806 stars30 Jun 2026: 843 stars7 Jul 2026: 873 stars14 Jul 2026: 915 stars21 Jul 2026: 957 stars28 Jul 2026: 992 stars4 Aug 2026: 1,028 stars11 Aug 2026: 1,073 stars18 Aug 2026: 1,112 stars25 Aug 2026: 1,161 stars1 Sept 2026: 1,367 stars8 Sept 2026: 1,510 stars15 Sept 2026: 1,672 stars22 Sept 2026: 1,780 stars29 Sept 2026: 1,856 stars6 Oct 2026: 1,985 stars13 Oct 2026: 2,281 stars20 Oct 2026: 2,461 stars27 Oct 2026: 2,855 stars3 Nov 2026: 3,247 stars10 Nov 2026: 3,752 stars17 Nov 2026: 4,417 stars

318 on 28 Apr 20264,417 on 17 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

Tauri-based desktop app (TypeScript) that puts chat, a coding agent, tool-using agents, and image and video generation in one window. It uses an existing Ollama or LM Studio engine or installs one, and installs and manages ComfyUI for image and video. Optional paid LU Cloud runs larger models on hosted GPUs; local mode is free.

Who it is for: Desktop users wanting local chat, image/video and coding agent, no terminal

Strengths

  • Installers for Windows, Linux and macOS, plus an Android APK; no Docker or config files
  • Installs and updates ComfyUI itself, with text-to-image, image-to-video and LoRA lanes
  • Coding agent shows diffs before applying; Local API is OpenAI-compatible behind a token
  • Agent tool calls pass a permission gate; supports your own MCP servers

Weaknesses

  • macOS build is Apple Silicon only, beta, and has run on one Mac so far
  • Windows installer is not yet code-signed, so SmartScreen warns on first run
  • Voice and Computer control are beta and mainly tested on Windows
  • Upscale, Erase Object and the largest models need the paid cloud tier

What it needs

  • RAM ≥ 8 GB
  • GPU optional
  • Needs Ollama or LM Studio (installed by the app if missing), ComfyUI (installed by the app)
  • Models: Ollama models, LM Studio models, GGUF, SDXL, FLUX

What changed

From the release notes. No flag means the note does not say so, not that the upgrade is safe.

Also in Desktop AI apps

See all 73
Also in Desktop AI apps
RankProjectScore
1cherry-studioDesktop client for chatting with many LLM providers, with agents and knowledge bases53.7k stars+1,096 (+2.1%) in 30 daysAGPL-3.093 out of 100
2screenpipeRecords screen and audio locally, makes it searchable for you and your agents22.5k stars+473 (+2.2%) in 30 dayscustom license92 out of 100
3dsh-desktopNative Windows and macOS desktop client for DeepSeek Harness, built on plugins36.4k stars+4,936 (+16%) in 30 daysMIT91 out of 100
#4hermes-desktopDesktop GUI for installing and chatting with Hermes Agent14.6k stars+264 (+1.8%) in 30 daysMIT88 out of 100
#5artcraftDesktop app for staging 2D and 3D scenes before generating images and video Live demo ↗ (opens in a new tab)41.6k starsApache-2.087 out of 100