Analysis of 2.3k Local AI Apps Reveals 82 Categories and Diverse Use Cases

An analysis of the Mac App Store identified 2,259 local AI applications out of over 20,000 scraped entries, highlighting a growing ecosystem of niche tools that package models with specific workflows. The survey covers 82 distinct categories, ranging from common tasks like transcription and OCR to specialized functions such as wardrobe styling and pet health assistance.

• Data was collected on June 24 by scraping the Mac App Store API using 513 search terms and developer profile crawling.
• A DeepSeek V4 Flash classifier filtered the results down to 2,259 apps confirmed as local AI.
• Categorization and grading utilized a combination of scripts and LLM-as-a-judge methodologies.
• The ranking system prioritizes applications that run entirely on-device.
• The resulting directory is available at bunnysoft.app/local-ai-mac-apps with plans for monthly updates.

This survey demonstrates that local AI apps are expanding beyond technical tinkering into durable, diverse professional and personal workflows, reducing the operational friction associated with command-line tools.