Hugging Face's Most Popular Models

Hugging Face continues to be a leading platform for sharing and discovering new artificial intelligence models. This week, several models have caught the community's attention, standing out for their number of downloads and "likes."

  • zai-org/GLM-4.7-Flash: This 31 billion parameter model is designed for fast and efficient text generation, ideal for real-time applications. It has totaled 124,000 downloads.
  • unsloth/GLM-4.7-Flash-GGUF: A quantized version of GLM-4.7-Flash, designed for local inference on systems with limited resources. It reached 112,000 downloads in a few hours.
  • zai-org/GLM-Image: A powerful model for generating and editing images from text, with 10,800 downloads.
  • google/translategemma-4b-it: Multimodal (vision and language) model with 5 billion parameters for translation and understanding of images and text.
  • kyutai/pocket-tts: A compact model for speech synthesis (TTS) that produces natural voices, suitable for mobile devices.
  • microsoft/VibeVoice-ASR: ASR (Automatic Speech Recognition) system with 9 billion parameters for low-latency multilingual speech recognition.
  • Lightricks/LTX-2: Model for creating professional-quality videos from images.
  • stepfun-ai/Step3-VL-10B: Multimodal model for advanced reasoning from images and text.