Hugging Models (@HuggingModels) header
Hugging Models (@HuggingModels) profile picture

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

Joined December 2025
People after checking the first reply on the original post.
“The way to build an agent is to ask an agent to build it for you.” — Jensen Huang So we stopped building Mercury's features and asked Mercury to build them. Backed by Open source ❤️
1
19
4,892
Meet OTel-2.0-Decision-26B-A4B-4bit: a 26B parameter decision engine for telecom. It classifies and makes typed decisions, bringing system-one thinking to network ops. Big news for open telecom AI.
2
26
2,613
Meet the uncensored Gemma 4 26B model that's making waves! This GGUF version is fully decensored and abliterated, perfect for unrestricted AI conversations. With 1.1k downloads already, it's a must-try for anyone exploring open models.
7
27
1
508
24,552
Name a biggest scam other than DUNS number?
3
1
5
2,133
Meet openamer-max-watch: a text-generation model built for German, Bosnian, and English. With 675 downloads and a local-first design, it's a hidden gem for multilingual creators who want control. Let's explore what it can do.
2
16
2,539
Meet DaneshSelwal/Classification-Hissar25-Models: a Keras-based image classification model ready to sort your visuals into categories. With 232 downloads, it's gaining traction in the community. Curious? Let's dive in!
2
2
9
2,498
Meet dale-0.8b-GGUF: a tiny but mighty AI that turns plain English into bash commands. Perfect for CLI lovers who want to speed up their workflow. Only 0.8B params, runs locally with llama.cpp. #NL2Bash #AI
4
2
17
2,537
Meet FLUX.3 Image Klein: a new image-to-image diffusion model that's blowing up on Hugging Face with 1.1M+ downloads. It's built for ComfyUI and ready to transform your images with stunning edits. Here's what makes it special. 🧵
2
10
1
130
9,422
Meet Arete-27B, a 27B text-generation model fine-tuned for value alignment. It's built on Qwen3.8-27B and trained on Arete data pairs. If you care about models that chat with purpose, this one's worth a look.
3
1
27
3,078
Meet DeepSeek-V4-Pro-Qwen3.5-9B-UNCENSORED-heretic: a 9B powerhouse blending Qwen3.5 and DeepSeek reasoning. With 27k downloads and 7.5k likes, it's turning heads for its uncensored, heretic approach to text and image-text tasks.
7
38
497
18,581