Hugging Models✓

This is insane: An agent can do what?

Introducing Optimus, the eyes and hands of Mercury Agent. Your AI agent can now see your screen, open apps, move the mouse, click, and type across macOS, Windows, and Linux, controlled through chat or voice.

Meet Ling-3.1-flash-1019: a new text-generation model with a hybrid architecture that's turning heads. It's fast, efficient, and ready for conversation. Here's why it matters for the AI community. 🧵

With 832 downloads and 80 likes, Ling-3.1-flash-1019 is gaining traction. Its fp8 optimization and hybrid design deliver top-tier performance for conversational tasks. Try it and see the difference.

🔗 Explore the full model details here:

Meet LFM3-1.7B-Base, a 1.7B parameter text generation model built for edge devices. It's lightweight, conversational, and supports English and Arabic. Perfect for on-device AI!

It's already got 1.4k downloads and 84 likes, proving its appeal. If you need a compact, multilingual text generator that runs locally, LFM3-1.7B-Base is a solid pick.

🔗 Explore the full model details here:

Meet DeepOpen Banking77 Sam13 v1, a ModernBERT-based text classifier fine-tuned for banking intent detection. Lightweight, fast, and ready to power your customer support automation.

With 222 likes and growing, it's a solid choice for research and production. Lightweight, accurate, and easy to deploy. Give it a try for your next banking NLP project!

🔗 Explore the full model details here:

Meet LiquidAI/d1-2B: a 2B parameter vision-language model built for edge devices. It classifies, extracts, and decides from images and text, all while running efficiently on limited hardware. Perfect for real-time AI on the go.

With 1,776 downloads and 91 likes, it's gaining traction for edge AI. Its liquid design enables adaptive computation, delivering reliable classification and extraction without cloud dependency. A solid pick for resource-constrained vision tasks.

🔗 Explore the full model details here: