Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
TCM (Truncated Consistency Models)github.com · Low evidenceTCM (Truncated Consistency Models) is an official implementation of a paper accepted at ICLR 2025 that introduces a method for training generative models with truncated consistency. The project provides pre-trained models and training/evaluation code for CIFAR-10 and ImageNet datasets, achieving low FID scores in few-step image generation.
Parse Score