Summary
LM Manager Pro is a local-first iOS and iPadOS app for managing
LM Studio inference servers and cloud AI providers. There are no
accounts, no analytics, no third-party crash reporting, and no
background telemetry. Everything you create — machines, personas,
chat history, episodes, scores — stays on your device unless you
deliberately submit content through the public feedback form.
What we store on your device
-
App data (SwiftData): machine configurations, AI personas,
teams, projects, chat sessions and messages, episodes, scoring
history, and prompt templates. Stored locally in the iOS sandbox
for this app only.
-
API keys (iOS Keychain): if you choose to use cloud
providers, your API keys for OpenAI, Anthropic, Google, and Qwen
are stored securely in the iOS Keychain. They are used only to
make API calls on your behalf and never sent anywhere else.
-
Settings (UserDefaults): non-sensitive preferences
such as theme, default sort order, and feature toggles.
What leaves your device
The app makes network requests only in these situations:
-
Local LM Studio servers: the app connects to machines you
explicitly add — over your local network or your Tailscale VPN —
to load models and run inference. These connections never touch
the public internet.
-
Cloud AI providers (only when you choose a cloud model):
prompts, attached images, and any chat context are sent to the
provider you selected. Each provider has its own privacy policy
that governs that data:
We do not relay, log, store, or proxy these requests — they go directly
from your device to the provider's API endpoint.
-
Web search (DuckDuckGo, optional, per persona): if you enable
web search for a persona, the app sends the persona's search query
to DuckDuckGo.
-
In-app feedback (optional): when you deliberately submit
feedback, the app sends the feedback type, title, details, app
version, build number, platform, and iOS or iPadOS version to a
developer-operated Cloudflare Worker. The Worker creates a
persistent public issue in the
LM Manager Pro Feedback repository.
The form states that submissions become public and warns you not to
include API keys, credentials, private hostnames or IP addresses,
proprietary prompts, or other sensitive information.
The developer does not receive your prompts, responses, provider
keys, chats, generated media, or local app database through the
model, media, or search features.
Voice features
Speech-to-text (dictation) and text-to-speech use Apple's on-device
speech and synthesis frameworks. Voice data does not leave your
device for these features.
Analytics, tracking, and ads
None. The app contains no analytics SDKs, no crash reporting,
no advertising identifiers, no third-party trackers, and no cookies.
The optional feedback relay is used only when you choose to submit
feedback; it is not analytics or tracking.
App privacy disclosure
Feedback text is treated as Other User Content. The attached app,
build, platform, and operating-system details are treated as Other
Diagnostic Data. This data is used for app functionality and
customer support, is not linked to an account or advertising profile,
and is not used for tracking.
Children's privacy
LM Manager Pro is rated 13+ because it can connect to AI chat and
content-generation services. The app does not knowingly collect
information from children. Users should not submit personal or
sensitive information through public feedback. User-configured cloud
providers have their own age requirements and privacy policies.
Your data, your device
You can export your full dataset as JSON for backup or migration to a
new device, and import it back with one tap. Deleting the app from
your device removes all locally stored data, including Keychain entries
created by the app.
Changes to this policy
If this policy ever changes, we'll update the date below and post the
new version at this URL. Material changes will also be noted in the
app's release notes.
Contact
Questions, concerns, or privacy reports?
Email Michael.d.williams13@icloud.com
or use the Feedback page to open a private email draft.
