AI Model / On-device / Edge LLM
An open 45M-parameter agentic LLM for tiny devices, built for tool calling, device use, and structured extraction. It ships as a 14 MB binary and is positioned for phones, wearables, smart home devices, robots, and other low-RAM hardware.
The page clearly presents a real AI product with a specific use case: an on-device agentic model for tool calling, device control, and structured extraction. It includes concrete evidence of size, RAM use, supported device classes, performance claims, licensing, and a playground/demo, which makes it useful for a public listing.
The crawl does not show the full docs, model access flow, or installation steps in enough detail to tell how a builder would integrate it end to end from the page alone.
Last evaluated 20 days ago. Current rank #633. Down 3 spots in the rankings.
A 150M ModernBERT-based token classification model that extracts verbatim evidence spans for RAG and related pipelines, reducing the need for an LLM call to highlight answers.
JetBrains introduces Mellum2, an open-source 12B model built for software engineering workflows, routing, Q&A, RAG, sub-agents, and private deployment.
An open-source 26M parameter function-calling model designed to run on very small devices, with a local quickstart, Python usage, and a playground for testing and finetuning.
