A desktop AI studio·Chat · Code · Image · Video · Remote
WINDOWS · ~7 MB installer · AGPL-3.0
Your machine. Your models. Your rules.
What it does
Twenty-plus presets auto-detected: Ollama, LM Studio, vLLM, KoboldCpp, llama.cpp, OpenAI, Anthropic, Groq, OpenRouter, and eleven more. Day-0 Qwen 3.6, GPT-OSS, GLM-4.7 Flash, DeepSeek R1, Gemma 4, Mistral Small 3, Phi 4. Vision. Memory. RAG. Personas. Thinking-mode. A/B compare.
Codex with live token streaming and apply-patch. Claude Code CLI integrated. An Agent Mode with fourteen tools, parallel execution, MCP, and sub-agent delegation. No separate IDE, no context-switching.
ComfyUI runs in the background. Image: FLUX 2 Klein, FLUX.1, Juggernaut XL, Z-Image Turbo, ERNIE-Image, SDXL, SD 3.5. Video: Wan 2.1, HunyuanVideo 1.5, LTX 2.3, FramePack F1, AnimateDiff. Image-to-image and image-to-video. Seventy-five-plus one-click downloads, hardware-aware.
LAN or Cloudflare Tunnel. Six-digit passcode. QR code setup. Full mobile web app: chat, Codex, tools, plugins. Your desktop does the compute — your phone drives it from anywhere.
Why
Most local-AI tools stop at text chat. This one gives you the whole studio: text, code, images, video, and your phone can drive it from anywhere. No data leaves your box unless you tell it to. No subscription. No telemetry.
Open source under AGPL-3.0. Auto-detects twelve backends and downloads models in one click. Signed auto-update channel, covered by 2,200+ unit tests.
Setup
No Docker. No terminal. No config files. Run the installer, let the wizard scan your system, start chatting.
One installer. Windows · Linux. Auto-updates over a signed channel. AGPL-3.0.
First launch scans twelve local backends — Ollama, LM Studio, vLLM, KoboldCpp, Jan, llama.cpp, and more. One-click install links if nothing is running.
Pick a model, start chatting. Switch to Codex for coding, open Create for images or video. Add cloud providers in Settings any time.
Models
Auto-detects models from any running backend. Day-0 support for Qwen 3.6, GPT-OSS, GLM-4.7 Flash, DeepSeek R1, Llama 4, Gemma 4, Mistral Small 3, and Phi 4. Seventy-five-plus one-click downloads, hardware-aware recommendations, VRAM-tier filtering.
Google flagship. Native tools, vision, Apache 2.0. E4B runs on 4 GB, 27B on 16 GB.
Strongest reasoning and coding. Qwen 3.6 (35B MoE, vision, 256K context, day-0). GPT-OSS-120B / 20B via Ollama. GLM-4.7 Flash. DeepSeek R1 and Llama 4 ready.
Popular SDXL finetune. Strong photorealism, community favourite. Text-to-Image and Image-to-Image. 8-10 GB VRAM.
Explicitly uncensored. Eight to fifteen seconds per image. No safety filters. T2I and I2I.
Lightricks LTX-Video. Fast text-to-video on modest hardware. Long clips, sharp motion, one-click setup.
Image-to-video on just 6 GB VRAM. Upload an image, get video. Next-frame prediction.
Writing
Everything worth knowing about running AI locally — hardware picks, which models matter, how the tools stack up against the competition.
27B dense, 35B MoE, NVFP4, BF16. Hardware picks, GGUF links, one-click install.
GuideQwen 3.6, Gemma 4 Heretic, Llama 3.1, Hermes 3. What abliteration is, where to download.
ReleaseSingle-instance lock, configurable HuggingFace path, in-app Privacy section, Linux drag fix.
GuideAll sizes from E4B to 27B. Native tools, vision, uncensored variants.
GuideUpload a photo, adjust denoise, transform. FLUX, Z-Image, SDXL.
ComparisonComplete comparison of GPT4All, Open WebUI, LM Studio, Jan, and more.
GuideSetup guide. Models, hardware, and why local beats cloud.
VersusBoth open source. Only one does chat + code + images + video.
VersusOpen source all-in-one vs polished closed-source chat client.
GuideHow LU handles ComfyUI setup, models, and workflows automatically.
GuideWan 2.1, HunyuanVideo, LTX, FramePack. Hardware requirements, model picks.
Questions
A free, open-source desktop app for running AI locally. Combines chat (20+ provider presets), a coding agent (Codex) with fourteen tools, image generation via ComfyUI (FLUX 2, Juggernaut XL, Z-Image, SDXL), and video generation (Wan 2.1, HunyuanVideo, LTX 2.3, FramePack F1) in one interface. AGPL-3.0 licensed.
Yes. Qwen 3.6 has day-0 integration (35B MoE, vision, agentic coding, 256K context). GPT-OSS-120B and GPT-OSS-20B run via Ollama. GLM-4.7 Flash supported through Ollama. Also Day-0 ready: DeepSeek R1, Llama 4, Gemma 4, Mistral Small 3, Phi 4.
Yes. Locally Uncensored works as a ChatGPT and Claude alternative that runs on your own hardware. Use Qwen 3.6, GPT-OSS, GLM-4.7, DeepSeek R1, Llama 4 or Gemma 4 instead — or add cloud providers (OpenAI, Anthropic, OpenRouter, Groq) alongside the local stack.
Yes. After setup and model download, no internet is needed for the local providers. No accounts, no telemetry, no usage limits. Cloud providers are optional — the core runs one-hundred percent on your hardware.
Those tools handle text chat. Locally Uncensored adds a coding agent with fourteen MCP tools, image generation, video creation, A/B model comparison, local benchmarking, granular permissions, file upload with vision, and thinking mode — all in one app.
Text chat: 8 GB RAM. Image generation: NVIDIA GPU with 8+ GB VRAM. Video generation: 10-12 GB VRAM. The app auto-detects hardware and recommends models. Windows 10/11 and Linux supported.
Abliterated models with artificial restrictions removed. The AI responds honestly without refusing or adding disclaimers. Combined with local execution, your conversations stay private.
Only if you explicitly dispatch a chat over LAN or Cloudflare Tunnel. Remote is opt-in, gated behind a six-digit passcode, and you see exactly when a device is connected. No background uploads. No telemetry.
Not yet. Windows and Linux for now. macOS support is on the roadmap but not promised. The source is AGPL-3.0 if you want to build for your platform.
WINDOWS · ~7 MB installer · AGPL-3.0