
Private LLM - Local AI Chat 13+
Local Offline Private AI Chat
Numen Technologies Limited
-
- $9.99
Screenshots
Description
Meet Private LLM: Your Secure, Offline AI Assistant for macOS
Private LLM brings advanced AI capabilities directly to your iPhone, iPad, and Mac—all while keeping your data private and offline. With a one-time purchase and no subscriptions, you get a personal AI assistant that works entirely on your device.
Key Features:
- Local AI Functionality: Interact with a sophisticated AI chatbot without needing an internet connection. Your conversations stay on your device, ensuring complete privacy.
- Wide Range of AI Models: Choose from various open-source LLM models like Llama 3.2, Llama 3.1, Google Gemma 2, Microsoft Phi-3, Mistral 7B, and StableLM 3B. Each model is optimized for iOS and macOS hardware using advanced OmniQuant quantization, which offers superior performance compared to traditional RTN quantization methods.
- Siri and Shortcuts Integration: Create AI-driven workflows without writing code. Use Siri commands and Apple Shortcuts to enhance productivity in tasks like text parsing and generation.
- No Subscriptions or Logins: Enjoy full access with a single purchase. No need for subscriptions, accounts, or API keys. Plus, with Family Sharing, up to six family members can use the app.
- AI Language Services on macOS: Utilize AI-powered tools for grammar correction, summarization, and more across various macOS applications in multiple languages.
- Superior Performance with OmniQuant: Benefit from the advanced OmniQuant quantization process, which preserves the model's weight distribution for faster and more accurate responses, outperforming apps that use standard quantization techniques.
Supported Model Families:
- DeepSeek R1 Distill based models
- Phi-4 14B model
- Llama 3.3 70B based models
- Llama 3.2 based models
- Llama 3.1 based models
- Llama 3.0 based models
- Google Gemma 2 based models
- Qwen 2.5 based models (0.5B to 32B)
- Qwen 2.5 Coder based models (0.5B to 32B)
- Google Gemma 3 1B based models
- Solar 10.7B based models
- Yi 34B based models
For a full list of supported models, including detailed specifications, please visit privatellm.app/models.
Private LLM is a better alternative to generic llama.cpp and MLX wrappers apps like Enchanted, Ollama, LLM Farm, LM Studio, RecurseChat, etc on three fronts:
1. Private LLM uses a significantly faster mlc-llm based inference engine.
2. All models in Private LLM are quantised using the state of the art OmniQuant quantization algorithm, while competing apps use naive round-to-nearest quantization.
3. Private LLM is a fully native app built using C++, Metal and Swift, while many of the competing apps are bloated and non-native Electron JS based apps.
Please note that Private LLM only supports inference with text based LLMs.
Private LLM has been specifically optimized for Apple Silicon Macs.Private LLM for macOS delivers the best performance on Macs equipped with the Apple M1 or newer chips. Users on older Intel Macs without eGPUs may experience reduced performance. Please note that although the app nominally works on Intel Macs, we've stopped adding support for new models on Intel Macs due to performance issues associated with Intel hardware.
What’s New
Version 1.9.11
- Support for two Qwen3 4B Instruct 2507 based models: Qwen3 4B Instruct 2507 abliterated and Josiefied Qwen3 4B Instruct 2507 (on Apple Silicon Macs with 16GB or more RAM)
- Fix for the rare crash in the Settings panel on some Macs.
- Minor bug fixes and updates
Ratings and Reviews
Good for the end of Shortcuts
Really interesting app! Thank you!
Just note that like in the various sample shortcuts, it’s best used at the end of your shortcut and then drop the result to the clipboard. I could never convince it to open another app afterwards and put the result in there directly.
Intriguing but
Writes excellent English and responds in roughly the same time as it would take to manually type the response. A long input, eg re-write this text (a whole web page), will often have no response at all.
I don’t know how it manages to work offline but it seems to work in airplane mode.
Generally the answers seem a bit limited and even more factually suspect than with ChatGPT and it will often insist on the same answer after being corrected. (The PM of NZ is Jacinda Ardern)
I’m hoping that with Shortcuts support coming it could be good for writing prettier versions of rough text.
Developer Response ,
Thanks for the feedback! The app currently contains a quantised 3B parameter decoder-only (aka GPT) LLM that runs on your device, and the app makes no network connections, whatsoever. This is how it works in Airplane mode, while apps like ChatGPT cannot. This is also the reason why the app is a fairly large download (1.6GB), even with data compression.
The context length of the current model in the app is 2048 tokens or about 1500 words (a token roughly corresponds to ~0.75 words). The context length of an LLM model is the most amount of text that the model can attend to. For comparison, the baseline GPT-3.5 and GPT-4 models that ChatGPT uses have context lengths of 4096 and 8192 tokens, respectively. In some ways it isn't a fair comparison, because the former (our app) is an app runs on your iPhone, while the other needs multiple large servers, and an active internet connection from your phone to those servers, to run. Anyway, improvements to the context length within the realm of possibilities and an active area of research. I can't promise anything but we might be able to increase the context length by a bit, soon. We're also experimenting with newer, larger models which have longer context lengths; although, they'll only work on newer iPhones and iPads.
WRT the model's factual knowledge, since the app doesn't connect to the internet, its knowledge is limited to what the model's training data contained. Incidentally, I asked ChatGPT (3.5) the same question question, and it came up with the same incorrect response that you noted.
Shortcuts integration will indeed ship later this week, we hope you'll like it!
App Privacy
The developer, Numen Technologies Limited, indicated that the app’s privacy practices may include handling of data as described below. For more information, see the developer's privacy policy.
Data Not Collected
The developer does not collect any data from this app.
Privacy practices may vary, for example, based on the features you use or your age. Learn More
Information
- Seller
- Numen Technologies Limited
- Size
- 1.3 GB
- Category
- Utilities
- Compatibility
-
- iPhone
- Requires iOS 17.0 or later and a device with the A12 Bionic chip or later.
- iPad
- Requires iPadOS 17.0 or later and a device with the A12 Bionic chip or later.
- Mac
- Requires macOS 14.0 or later.
- Apple Vision
- Requires visionOS 1.0 or later and a device with the A12 Bionic chip or later.
- Languages
-
English
- Age Rating
- Learn More
- Infrequent Mature or Suggestive Themes Horror/Fear Themes Medical Treatment Information
- Copyright
- © 2024 Numen Technologies Limited
- Price
- $9.99
Supports
-
Family Sharing
Up to six family members can use this app with Family Sharing enabled.