Chat and instruct modes
Instruction-following, character chat, message editing, and conversation branching.
Your AI, on your machine, fully private
Chat with open AI models — 100% offline and private.
Download, unzip, double-click. Load a GGUF model and talk to it with chat, vision, tool-calling, and an OpenAI-compatible API — with zero telemetry.
Desktop app built
Windows · macOS · Linux
No charge today. We will share final pricing, platform support, and delivery terms before you decide to buy.
CAW TextGenProduct screenshots coming soon
Zero telemetryNo external resources or remote update requests
Open sourceAGPL-3.0 fork of oobabooga/textgen
Your hardwareCUDA, Vulkan, ROCm, and CPU-only builds
Focused workflow
Everything you need to move the work forward
A practical tool for a specific operational job, with less setup and less clutter.
Vision
Attach images to messages for visual understanding.
Tool-calling
Models can call custom functions — web search, page fetching, math — plus MCP server support.
OpenAI-compatible API
Chat, Completions, and Messages endpoints: a local drop-in for the OpenAI/Anthropic APIs.
Training and image generation
Fine-tune LoRAs and generate images with diffusers models in the full installation.
Extensions
TTS, voice input, translation, and community extensions.
How it works
From task to trusted record in three steps
Download
Get the portable build for your OS and GPU.
Add a model
Drop a GGUF file from Hugging Face into the models folder.
Chat
Open the app and talk — everything stays on your machine.
Release status
Clear about what is ready
We are preparing checkout, signing, licensing, and downloads. The page will switch to purchase actions only after the customer artifacts are verified.
Desktop app
Desktop app builtCAW TextGen is a built, rebranded fork of oobabooga/textgen (AGPL-3.0 — the fork's source stays open). Portable builds cover Linux, Windows, and macOS with CUDA, Vulkan, ROCm, and CPU-only options.
Android app
Windows · macOS · LinuxCAW TextGen is a desktop app. No mobile version is planned.
Launch note
Coming soonLicensed AGPL-3.0 like the upstream project. If you distribute or host it, the source must stay open.
Product questions
What teams ask before they reserve
Is CAW TextGen really private?
Yes. It runs 100% offline with zero telemetry, no external resources, and no remote update requests. Your chats never leave your machine.
What is CAW TextGen based on?
It is a Custom AI Works-branded fork of the open-source text-generation-webui by oobabooga, licensed AGPL-3.0. The original work remains the property of its contributors.
Need this workflow in your business?
This app is in development. We can still scope a workflow around your team, systems, and requirements.