Choose an app for the job

Choose a desktop app for simple chat, a document workspace for questions about files, or a model runner to connect AI to other programs.

Choose by job

Start with what you want the app to do.

A desktop app is usually the easiest first step. Choose a document workspace for file questions, or a model runner when another program needs to use your model.

Ollama or LM Studio?

Considering other apps? Compare apps by task, platform, and privacy settings.

Search softwareJob

Desktop apps for chat

Browse, download, and chat with models in one app.

Desktop app

LM Studio

A straightforward way to find, download, and chat with a model in a desktop app.

Use it when: you want to browse, download, and chat with models in a desktop app.
Avoid it when: you need a shared browser workspace with user accounts; pair its server with a suitable interface instead.
Where your data goes: Downloaded models and runtimes support offline chat, document chat and the local server. Discovery and downloads need internet. Cloud models, web search, LM Link and connected tools add other destinations; leave those features unused for work that must stay on this computer.
Version covered: 0.4.26

See setup advice

Desktop app

Jan

A desktop alternative when you want local chat plus optional provider connections in one app.

Use it when: you want a desktop chat app with optional connections to other model providers.
Avoid it when: you need a centrally managed browser workspace for several users.
Where your data goes: Downloaded local models run on your computer. Cloud providers receive your messages; web search and fetch tools also contact external services. Leave those tools off for offline work. Choose product analytics preferences at first launch or change them in Settings.
Version covered: 0.8.6

See setup advice

Desktop app

GPT4All

Consider it for a desktop chat and LocalDocs workflow, but compare the current product state against newer alternatives first.

Use it when: you want desktop chat and questions about files through LocalDocs.
Avoid it when: your required model format or hardware backend is not supported by the app.
Where your data goes: Downloaded local models run on your device. LocalDocs creates embeddings from a selected folder and supplies matching passages to the selected chat model. Keep that model local, review analytics and data-sharing settings, and keep private source folders outside cloud sync. Initial downloads need internet.

See setup advice

Run models for other apps

Run a model on your computer and let another program use it.

Run models for other apps

Ollama

Desktop chat on Windows and Mac, terminal commands, and a local API for other apps.

Use it when: you want desktop chat on Windows or Mac, terminal commands, or a local API for other apps.
Avoid it when: you want a complete document workspace without connecting another app.
Where your data goes: Downloaded local models process prompts on this computer. Cloud models and web search send requests outside it. Set OLLAMA_NO_CLOUD=1 and restart for local-only mode; the log should say Ollama cloud disabled: true. Keep the default loopback address unless other devices need access. Downloads still need internet.
Version covered: 0.40.6

See setup advice

Browser interfaces and shared workspaces

Use a browser to chat with models from the providers you choose.

Browser workspace

Open WebUI

Use it as a browser interface after the underlying model provider already works.

Use it when: you want a browser chat workspace connected to a model that already works.
Avoid it when: your model provider is not working yet or you want the fewest moving parts.
Where your data goes: Your browser connects to the Open WebUI server, which contacts your selected model and document-processing services. Choose local inference, extraction, embeddings and reranking in Admin Settings for an on-device workflow. Offline mode controls selected downloads and checks; it is not a firewall. External search, tools and hosted providers can send content elsewhere.
Version covered: 0.12.3

See setup advice

Search and chat with documents

Best when you want to import documents, search their contents, and ask questions about them.

Document search and chat

AnythingLLM

Organize documents, search their contents, and ask questions about them.

Use it when: you want to organize files and ask questions across your documents.
Avoid it when: you only need simple chat and do not need document search or workspaces.
Where your data goes: Use local chat and embedding providers and a local vector database for a single-computer document workflow. Anonymous telemetry can be disabled. Removing a document from a workspace does not delete it from My Documents; delete it there to remove the parsed source and cached embeddings. Exports and backups need separate deletion.
Version covered: 1.18.0

See setup advice

Advanced model engines

For experienced users who want to control how models run and connect the engine to other software.

Advanced model engine

llama.cpp

For experienced users who want detailed control over model files, processor settings, and connections to other apps.

Use it when: you are comfortable with commands and need detailed control over how a model runs.
Avoid it when: you want a beginner desktop experience rather than detailed runtime controls.
Where your data goes: A downloaded GGUF runs locally without a cloud inference provider. The -hf option contacts Hugging Face for downloads. Server mode accepts prompts from other apps; its listen address determines who can connect. Connected tools have their own network access. Download first and run the local file for an offline test.
Version covered: rolling repository

See setup advice