Hugging Models

We're sharing/showcasing best of @huggingface models. Follow to stay in loop. Promoting Open-Source models.

[Hugging Models @HuggingModels]

🚨Just in: An open-source AI that uses your computer like a person, and calls you on WhatsApp when it's done.

Press a key. Talk. Watch it work.

Your AI agent just got hands.

Optimus sees your screen, moves the mouse, types like a human, and stops the second you say "stop."

[Mercury @mercury__agent]

Released Mercury 1.4.0, "Pluggy Mercury" ☿

Plugins are here. The first 2:

• WhatsApp: chat, or talk on a voice call. It can even call you.
• Optimus: Mercury uses your computer. It sees the screen, moves the mouse & types like a person.

Open source. macOS, Windows, Linux.

[Hugging Models @HuggingModels]

🔗 Explore the full model details here:

https://t.co/q1mDjDzNgm

[Hugging Models @HuggingModels]

Meet Muse-Spark-Open-816B-A42B by meta-models: a 816B parameter image-text-to-text model that's sparking conversations. With 5k+ downloads and 673 likes, it's the talk of the town. Ready to explore?

[Hugging Models @HuggingModels]

The community loves it: 673 likes and 5k downloads show its value. It's a powerful tool for multimodal tasks, and with the Muse-Spark lineage, you know it's built for quality. Dive in and see what you can create!

[Hugging Models @HuggingModels]

🔗 Explore the full model details here:

https://t.co/kp4wMoiDsX

[Hugging Models @HuggingModels]

Meet FLUX.3 Image Klein: a ComfyUI-ready image-to-image model that edits and transforms your images with stunning precision. Over 534k downloads and 1.3k likes prove the community is loving it. Ready to see what it can do?

[Hugging Models @HuggingModels]

With 534k+ downloads, it's a proven favorite. Expect fast, high-quality edits that keep the essence of your original image. Whether you're a hobbyist or a pro, this model delivers top-tier results every time.

[Hugging Models @HuggingModels]

🔗 Explore the full model details here:

https://t.co/CaoeE4IcCn

[Hugging Models @HuggingModels]

Meet Huihui-OmniCoder-9B-abliterated-GGUF: a 9B coding model built on Qwen3.5, now uncensored and ready for agent workflows. If you want a local code assistant without guardrail friction, this is worth a look.

[Hugging Models @HuggingModels]

The value is simple: a capable 9B coder that runs locally, handles agent tasks, and won't lecture you. With 807 downloads and growing, it's a practical pick for developers who want control over their stack. Try it for your next coding agent.

[Hugging Models @HuggingModels]

🔗 Explore the full model details here:

https://t.co/ANefTcChP7

[Hugging Models @HuggingModels]

Meet Flint-APEX-I-NanoPlus-V2.1-GGUF: a quantized text-generation model built for llama.cpp. With APEX quantization and NanoPlus architecture, it's a lean, efficient option for local AI. Perfect for tinkerers and devs who want performance without the bulk.

[Hugging Models @HuggingModels]

With custom quantization and the Flint tag, this model balances size and quality. It's a niche pick with 176 downloads, but for those needing a compact, capable text generator, it delivers solid performance.