Your session has been flagged for unusual activity.
Please confirm you're human to continue.
Local AI desktop app · Mac & Windows
Give your local model a desktop home. Connect Ollama or LM Studio to MultipleChat, run AI on your own hardware, and keep your conversations stored on your computer.
24-hour trial · $250 one-time license
Cloud API usage billed separately. Local models use your hardware.
Your computer Local model setupOne computer. App and model connected locally.
Illustrative setup. Model downloaded separately.
Take a closer look
This illustrated walkthrough introduces installation and your first conversation on Mac. Follow the local-model setup steps below to connect AI running on your own computer.
From download to your first conversation · Illustrated walkthrough; screens are simplified.
AI on your own hardware
MultipleChat stores your conversations on your computer. Connect a model running on that same machine to process your messages locally as well.
MultipleChat provides your chat interface, workspaces and local history. It connects to the model you select in Settings → AI Providers → LLM.
Ollama or LM Studio runs a downloaded model. MultipleChat sends it your message through a local connection and displays the answer.
| Part of your setup | How it works locally |
|---|---|
| Chat history | Stored by the desktop app on your computer. |
| Model responses | Processed on your machine when you choose a model running there. |
| Document indexing | Use a local embedding model as well to keep indexing local. |
| Internet access | Needed for initial downloads, cloud features and the licensed app's check-in at least every 14 days. |
A cloud provider or a model running on another computer changes where requests go. See how cloud models work with your own API keys.
16 GB RAM or more is a useful starting point. Model size, conversation length and other running apps all affect performance. Allow extra storage for model downloads, and test your chosen model during the trial.
Getting started
MultipleChat is the chat app. Ollama or LM Studio runs the model. Set up both on the same computer for local inference.
Have your API key or local model ready before clicking Start trial. Your 24 hours begin at that point. A consumer AI subscription does not automatically include API credit.
Choose the installer for your Mac or Windows PC. Leave the trial unstarted while you prepare your local model.
Install Ollama or LM Studio and download a model that fits your computer's memory. Make sure the model's local service is running.
In MultipleChat, open Settings → AI Providers → LLM. Select Ollama or LM Studio, use the local connection details from your model runner, and select the model you downloaded.
Start your trial, create a workspace and use the local model. Keep the model runner open. If you use document search too, configure local embeddings so that indexing also stays on your computer.
Your desktop app license
Try MultipleChat AI Desktop on your own computer before buying. Your license key arrives by email.
No recurring app subscription. Cloud model usage is billed separately by your provider.
Before you download
Local storage means the app saves your chats and files on your computer. Local AI means the model also processes your requests there. A cloud model can still be used by a desktop app with local storage; those requests go to the cloud provider.
You prepare the local model separately through Ollama or LM Studio, then connect it in MultipleChat. The model download needs additional storage, and the model runner needs to be running when you chat.
16 GB RAM or more is recommended as a starting point. The model size, conversation length and other running apps affect memory needs and speed. Check your chosen model's requirements and use the trial to test it on your hardware.
After downloading and setting up your models, local chat can work offline. Use local embeddings as well if you want local document indexing. Cloud models and web search need internet access, and a licensed copy of the app needs an online check-in at least every 14 days.
Local inference has no cloud provider fee per token. It uses your own computer's memory, storage and electricity. The MultipleChat desktop license costs $250 once after the 24-hour trial; the model's own license also applies.
Create an API key with a supported provider and add it in Settings → AI Providers → LLM. The app connects directly to that provider using your key. API usage is billed to your provider account separately from the app license.
Your chat history, documents, local document index and API keys are stored on your computer. When you choose a cloud model, your prompt and any relevant document passages are sent to that provider under its terms. No MultipleChat server sits between your app and the model provider.
Yes. Connect a local model through Ollama or LM Studio. Local chat and document indexing can run on your computer without cloud API usage charges. Model performance depends on your hardware. A licensed copy needs an online check-in at least every 14 days; web search and cloud models also need an internet connection.
Try the app for 24 hours. The trial starts when you click Start trial, so prepare your API key or local model first. A license costs $250 as a one-time purchase. Cloud API usage is charged separately by your provider. Local models use your own hardware without per-token provider fees.
No. MultipleChat AI Desktop is a separate, single-user product using your own API keys or local models. The web subscription and web-only features, including the Studios and AI Collaboration, are separate.
Your license key is emailed to the address you use at checkout. Enter it under Settings → License. If the email does not arrive, check your spam folder or contact [email protected].
Need help connecting your AI? Read the setup guide or contact support.
Your providers. Your desktop.
Connect OpenRouter or a direct provider account. Each guide shows which key to use, how to choose a model, and where your requests go.