Chatd.ai
Chatd.ai is a private, fully local desktop AI that chats with your documents offline via Ollama.
Last verified:
What is Chatd.ai?
Chatd.ai is a private, completely local desktop app that lets you chat with your documents using AI. All data stays on your computer and is never sent to the cloud. Chatd.ai ships with a built-in local LLM runner (Ollama with Mistral-7B), so there is nothing else to install — download, unzip, and run.
Chatd.ai uses RAG (Retrieval Augmented Generation) to index document content — including PDFs — and answer questions from your files instantly, fully offline. It is ideal for researchers summarizing lengthy papers, legal professionals, students, and anyone who works with documents and values privacy. Available for Mac, Windows, and Linux.
Chatd.ai pricing
Pricing model: Free
Chatd.ai is currently free. The application is available as a free download from chatd.ai or the GitHub releases page. No paid subscription plans are mentioned at this time.
Chatd.ai pros
- Privacy-focused: all data stays on your computer
- No cloud uploads or privacy concerns
- Comes with built-in LLM runner packaged in
- No extra installation needed—just run the executable
- Works offline without internet connection
- Intuitive and simple interface
- Time-saving: quickly extract information from documents
- Powered by Mistral-7B large language model
- Accurate and relevant responses
- Supports various file types including PDFs
- Uses Ollama for cross-platform LLM runner API
- Automatically manages AI model download and caching
- Open-source with MIT license
- 1.1k+ stars on GitHub
- Available for Mac, Windows, and Linux
Chatd.ai cons
- Local resource intensive: may require powerful computer
- Limited file format support for obscure types
- Windows app is not signed (shows warning when running)
- MacOS app needs developer certificate to sign for other systems
- GPU support requires advanced setup
- Custom model selection requires manual Ollama setup
- Only supports local LLM (no cloud API options)
- May have slower performance on older computers
Frequently asked questions about Chatd.ai
What is Chatd?
Chatd is a desktop application that lets you use a local large language model (Mistral-7B) to chat with your documents. All your data stays on your computer and is never sent to the cloud, making it a completely private and secure way to interact with your documents.
Do I need to install anything else to use Chatd?
No, you don't need to install anything else. Chatd comes with the local LLM runner (Ollama) packaged in, so you just download the latest release, unzip it, and run the chatd executable.
What file types does Chatd support?
Chatd supports various file types including PDFs. The app uses RAG to clean up PDFs (removing new lines, hyperlinks, images, citation numbers) and split content for embedding and vector storage.
Does Chatd work without an internet connection?
Yes, Chatd works offline without an internet connection. All data processing happens directly on your computer using local AI, ensuring maximum privacy and security.
What AI model does Chatd use?
Chatd uses the Mistral-7B large language model by default. It uses Ollama to run the LLM, and if you already have an Ollama instance running locally, Chatd will automatically use it.
Can I use a custom model with Chatd?
Yes, you can use a custom model. You need to download Ollama for your platform, pull the desired model from the Ollama library using 'ollama pull <model_name>', then open Chatd settings and type the model name in the Model Name field and save.
How do I download Chatd?
Download the latest release from chatd.ai or from the GitHub releases page. Then unzip the downloaded file and run the chatd executable.
Is Chatd available on all platforms?
Yes, Chatd is available for MacOS, Windows, and Linux. Each platform has specific packaging requirements, and you need to download the appropriate Ollama binary for your platform.
How does Chatd find answers in my documents?
Chatd uses RAG (Retrieval Augmented Generation). It fetches PDF data, cleans it up, splits on periods, generates embeddings for each sentence using 'Xenova/all-MiniLM-L6-v2', stores vectors in the Orama vector DB, then when you chat it converts your prompt to embedding, finds 20 near vectors, joins them to max 500 characters, and puts them into your prompt.
What is Chatd's license?
Chatd is licensed under the MIT license, making it open-source and freely available for use and modification.