About this app
PocketLM is a local AI app built for Apple devices, designed to make on-device inference faster, lighter, and more practical for everyday use. It supports both MLX and GGUF model types, with broad coverage across model series including Qwen3.5, Qwen3, LFM2.5, LFM2, MiniCPM5, Gemma4, Gemma3n, Gemma3, BitNet, ERNIE 4.5, and EXAONE 4. Why PocketLM? Local-first inference: core chat and reasoning tasks can run directly on your device. Dual format support: works with both MLX and GGUF models for greater flexibility. Broad model coverage: from compact text models to multimodal models with vision capability. Optimized for Apple Silicon: tuned for responsive performance on Apple hardware. More than chat: ask questions, write, summarize, reason, and understand images in one app. If you have any questions, please feel free to contact us at jiangdi0924@gmail.com. In-App Purchase: Payment will be charged to your Apple ID account at the confirmation of purchase. Terms: https://joychat.notion.site/Terms-7ae7c25789c14aa2a1691b5821dd882a Privacy: https://joychat.notion.site/Privacy-Policy-9b295e4625a94a2892ac457153880f82
Latest release
What’s New
Added support for the MiniCPM5 1B 4-bit model, including downloads from Hugging Face and ModelScope.
Audience growth
Ratings growth
Baseline captured. Growth appears after the crawler observes a changed rating count.
Momentum comes from crawler snapshots for the tracked storefront. “Viral growth signal” means unusually fast recent rating gains; it is an indicator, not download data.
Price tracking
Price intelligence
First snapshot captured. The trend line starts after the next crawler price observation.
- First observed Aug 8, 2026 · 1:40 PM Free
Screenshots
