Private LLM
Verified Oct 16, 2026 ActiveRuns 140+ open-weight chat models on iPhone, iPad and Mac with no account and no cloud call
Listed on EffectHub ·
Pricing
- Pricing model
- Paid
- Paid options from
- $4.99 once
- Billing
- One-time purchase
One-time App Store purchase at $4.99 in the US store, unlocking the app on iPhone, iPad and Mac with Family Sharing for up to six people. No subscription and no per-token cost, since inference runs on the device.
Use tool ↗About Private LLM
Private LLM runs language models locally on Apple hardware. Nothing leaves the device, there is no login, and inference makes no network call; the app ships quantised builds of open-weight families including Llama, Qwen, Gemma, Phi, Mistral and DeepSeek distills, with the catalogue put at over 140 models. On macOS it plugs into system-wide text editing, and across platforms it works through Siri, Shortcuts and the x-callback-url scheme for automation. It is sold as a one-time App Store purchase with Family Sharing for six rather than a subscription. The App Store seller is Numen Technologies Limited. The harvested address personalgpt.app now redirects here after a rename.
- API
- Not confirmed
- Last verified
- Oct 16, 2026
Verified sources
- US App Store listing: Private LLM - Local AI Chat at $4.99, seller Numen Technologies Limited, first released 5 June 2023, version 1.9.16 from 14 September 2026, iOS 17.0 or later, 4.2 stars over 735 ratings (15 October 2026)apps.apple.com ↗
- Home page: 140+ local models including DeepSeek R1 Distill, Llama 3.3, Qwen3, Phi 4 and Gemma 3 with OmniQuant and GPTQ builds, Siri, Shortcuts and x-callback-url support, macOS text tools, one purchase across Apple devices with Family Sharing for six (15 October 2026)privatellm.app ↗
Is Private LLM legit?
We don’t label tools safe or unsafe. Below are the records we checked ourselves, each with the date we read it.
- Domain registered
- Aug 10, 2023 (3 years ago)Registry record (RDAP), registrar Namecheap Inc., looked up Sep 6, 2026
- TLS certificate
- Issued by Google Trust Services, valid until Dec 1, 2026Seen in our own connection to the site on Oct 16, 2026
- Last checked
- Oct 16, 2026The site answered
- On the market since
- 2023The vendor's own material, read by us
- Operated by
- Numen Technologies LimitedOur research, card verified Oct 15, 2026
Pros & cons of Private LLM
The whole pitch is that inference happens on your device: no account, no upload, works with the network off, paid once instead of monthly. The model catalogue is wide and tuned for Apple silicon, and the Shortcuts and system text-service hooks make it useful beyond a chat window. Local models are still meaningfully weaker than hosted frontier models, and this is Apple-only.
Pros
- Inference is on-device: no account, no upload, works in airplane mode
- One-time price with Family Sharing for six instead of a recurring fee
- Wide catalogue of quantised open-weight models tuned for Apple silicon
- Hooks into Siri, Shortcuts, x-callback-url and macOS system text services
Cons
- Apple platforms only, with no Windows, Linux, Android or web version
- 1.4 GB install plus model downloads, and older devices are excluded (iOS 17 and A12 or newer)
- Local models are noticeably weaker than hosted frontier models
- App Store rating sits at 4.1, with reviewers flagging missing chat history that the vendor says arrives in v2
Best for
- people who want a chat assistant that never sends text off the device
- Apple users tired of monthly assistant subscriptions
- anyone automating text tasks through Shortcuts or Siri
Not for
- users on Windows, Linux, Android or the web
- work that needs frontier-model quality or long context
Watch out: The old personalgpt.app address now 308-redirects here because the product was renamed, so older reviews and screenshots describe a different app.
Alternatives to Private LLM
Private LLM head-to-head
Side-by-side fact sheets against the competitors our research or their makers name — pricing, platforms, API and both verdicts on one page.