v2.2 · 2026-06-12
LM Studio
runtime · hybrid●○○beginner
LM Studio is a local AI desktop app and developer stack. It can be used as a chat app, model manager, document-chat app, and local server.
FEATURES
+OpenAI API: yes
+Document chat: built in
-Docker: not required
-Multi-user: no
+GPU support: yes
UI: desktop_gui
API_BEHAVIOR
Desktop app can expose local API/server modes; serving beyond localhost changes the risk boundary.
PLATFORMS
macwindowslinux
BEST_FIT
+Beginner desktop GUI
+General chat
+Document workflow exploration
NOT_FIT
-Multi-user self-hosted web workspace
-Assuming offline/private behavior without checking settings
RUNTIME_OVERVIEW
LM Studio is a local AI desktop app and developer stack. It can be used as a chat app, model manager, document-chat app, and local server.
Install the app, choose a small first model, start a local chat, and only then try document chat or the local server.
Local chat can stay local in the right configuration, but model search, downloads, updates, runtime downloads, remote features, and server binding matter.
BEST_FOR
·beginner GUI workflows;
·model browsing and downloads;
·local chat in a desktop app;
·document attachments and local document experiments;
·optional local API/server workflows.
NOT_GOOD_FOR
·the lightest possible backend runtime;
·multi-user self-hosted web interfaces;
·claims about performance without hardware tests;
·assuming offline/privacy behavior without checking settings.
PLATFORMS
·mac
·windows
·linux
PROPERTIES
beginnerFriendly: high
setupDifficulty: easy
LINKS
EVIDENCE
Official documentation reviewed with caveatsreviewed: 2026-05-24
Official documentation reviewed, with caveats
CAVEATS
·Offline use is documented after model files are acquired, not guaranteed for every workflow.
·Local AI Guide has not independently installed, benchmarked, or audited LM Studio.
·Remote MCPs, linked devices, provider choices, and network serving can change the privacy boundary.