David
David Drobnitza is an independent developer and creator of Locally Uncensored, a free, open-source local AI studio that brings chat, image and video generation, coding and agent tools into one desktop application. The project emphasizes running AI privately on the user's own hardware, without mandatory cloud use, Docker or a terminal.
At a Glance
- Privacy-conscious individual AI users
- Open-source and self-hosting communities
- Artists and other creative professionals using generative image/video/audio tools
- Developers needing local coding agents and model experimentation
- +1 more
AI Tools by David
(1)
Locally Uncensored
Free, open-source desktop AI studio combining local chat, image and video generation, a coding agent and agent mode, with an optional hosted cloud.
Discussions
No discussions yet
Be the first to start a discussion about David
Latest News
Locally Uncensored 3.0.8 released with five-service web search, background video rendering, multi-GPU support, model-memory planning, AMD/Intel GPU support and settings/storage improvements.
David published coverage of DeepSeek V4.1 Flash Uncensored and related LU Cloud testing and pricing.
Locally Uncensored 3.0.7 release: REA reverse-engineering plugin, trained characters across Create and chat, ComfyUI workflow import, expanded editing and Android app.
Locally Uncensored 3.0.6 release: beta voice, beta computer control, Create in chat, additional LU Cloud models and first Apple-Silicon Mac build.
Products & Services
Free AGPL-3.0 open-source desktop AI studio for Windows and Linux, with an Apple-Silicon macOS beta and an Android app. It combines local chat, document/RAG chat, image and video creation, voice, coding-agent and general-agent tools, model management, gallery and remote access in one application.
Optional hosted companion service run by the same creator through lu-labs.ai. It provides larger hosted chat models and cloud image, video, audio and related Create operations through the Locally Uncensored app and a browser interface.
A separate project identified on David's GitHub profile as another creation; it relates to his BIM/AI/open-source work.
Market Position
Locally Uncensored positions itself as an all-in-one local AI studio rather than a chat-only frontend: it combines chat, image generation, video generation, coding and agent workflows in one open-source application. Its stated differentiators versus Open WebUI, LM Studio, Jan, SillyTavern, GPT4All, KoboldCpp and Msty are the unified creative workflows, one-click local setup, ComfyUI-backed image/video support and privacy/offline operation; LU Cloud adds a hosted path when local hardware is insufficient.
Leadership
Founders
David Drobnitza
Independent developer working at the intersection of BIM, AI and open source; GitHub identifies him as the creator of Locally Uncensored and Plan2IFC. He publishes AI and local-model tutorials under the PurpleDoubleD/ David identity and is the person named as responsible for locallyuncensored.com and the desktop app in the site's legal notice.
Founding Story
David says he built Locally Uncensored to address the fragmented local-AI workflow: separate tools and terminal windows for Ollama/chat, ComfyUI/images, video, voice and model management. The initial vision was a single, easy-to-install desktop app in which local inference, privacy and creative generation work together without cloud dependence.
Business Model
Revenue Model
The desktop application is free and open source for local use. Revenue comes from optional LU Cloud hosted inference sold through monthly plans and non-expiring credit packs; images, video, API calls and metered chat draw from a shared credit wallet, while selected Flash chat models are included within plan allowances.
Pricing Tiers
900,000 credits/month; 11 Flash models without credits and all 48 supported chat models.
2,350,000 credits/month; 12 Flash models without credits, including GLM 5.3 Flash.
5,000,000 credits/month; 12 Flash models without credits and maximum monthly cloud volume.
Non-expiring cloud credit packs; exact pack amounts are listed on the LU Cloud pricing flow.
Target Markets
- Privacy-conscious individual AI users
- Open-source and self-hosting communities
- Artists and other creative professionals using generative image/video/audio tools
- Developers needing local coding agents and model experimentation
- Users without a sufficiently powerful GPU who need hosted inference
- Private/offline conversational AI and document question-answering
- Local image, video, audio and character generation for creators
- Coding assistance and repository maintenance
- Agentic research, web browsing, file work and computer control with user approval
- Running AI on consumer Windows, Linux, Apple-Silicon Mac and Android hardware
- Hosted inference for users without a suitable GPU or needing larger models