Brief-details: Modern BERT-style model with 149M parameters, trained on 2T tokens. Features RoPE embeddings, 8K context length, and efficient attention mechanisms.
BRIEF-DETAILS: Aya-expanse-32b is a large language model from CohereForAI with 32B parameters, part of the Aya model family focused on expansive language capabilities.
BRIEF-DETAILS: Stability AI's latest turbo-charged SD model optimized for fast inference while maintaining high-quality image generation capabilities
BRIEF-DETAILS: Optimized checkpoint of flux1-dev for ComfyUI users with <24GB VRAM, includes dual text encoders in single safetensor file
Brief-details: Specialized float8 quantized version of FLUX models, offering float8_e4m3fn and float8_e5m2 weight variants for optimized performance while maintaining quality
Brief-details: Stability AI's medium-sized text-to-image model, part of the SD3 family. Optimized balance between performance and resource requirements.
BRIEF-DETAILS: Llama-2-7b: Meta's 7B parameter language model, part of the Llama family. Open-source with improved performance and safety features.
BRIEF-DETAILS: 14B parameter LLM based on Qwen 2.5 architecture with enhanced reasoning, 128K context window, and multilingual support across 29 languages
Brief-details: Viper-Coder-v1.6-r999 is a 14B parameter coding-specialized LLM based on Qwen 2.5, offering superior code generation, debugging, and multi-language support with 128K context window
Brief-details: A 14B parameter LLM based on Qwen 2.5 architecture, optimized for reasoning and multilingual support with 128K context window. Excels in structured responses and chain-of-thought reasoning.
Brief-details: A 14B parameter LLM based on Qwen 2.5 architecture, optimized for reasoning and multilingual support with 128K context window. Scores 39.75% average on benchmarks.
Brief-details: Lightweight Microsoft Phi-4 variant optimized for instruction following with 128K context length, available in multiple GGUF quantizations for efficient deployment
Brief-details: Long-context LLM supporting 1M tokens, 7.61B parameters, optimized for both long and short tasks with RoPE architecture and sparse attention mechanisms.
Brief-details: DepthPro-hf is a state-of-the-art monocular depth estimation model by Apple that generates high-resolution depth maps with exceptional detail and metric accuracy at 0.3s per image.
Brief-details: Gemma-2-2b-it is Google's instruction-tuned 2B parameter language model, built for chat and instruction-following tasks, requiring license acceptance on Hugging Face.
Brief Details: ControlNet v1.1 - Advanced neural network architecture for controlled image generation and manipulation, developed by lllyasviel on HuggingFace.
Brief-details: AMD's open-source 3B parameter LLM trained on MI300X GPUs, outperforming existing open models with strong reasoning and instruction-following capabilities
BRIEF-DETAILS: 70B parameter Llama-based model, distilled from R1 1776, designed to provide uncensored, unbiased responses while maintaining strong reasoning capabilities
Brief-details: A specialized hallucination detection model based on ModernBERT with 8192-token context support, achieving 79.22% F1 score for RAG applications.
Brief Details: A state-of-the-art multilingual Mixture of Experts text embedding model supporting 100+ languages, featuring 305M parameters and flexible embedding dimensions from 256-768.
Brief-details: ACE_Plus is an advanced AI model for instruction-based image creation and editing, featuring portrait consistency, subject manipulation, and local editing capabilities through context-aware content filling.
