GLM-Image
GLM-Image is a next-generation, open source image generation model developed by Z.ai, designed to combine deep language understanding with high-fidelity visual synthesis. Unlike traditional diffusion-only models, it uses a hybrid architecture that integrates an autoregressive language model with a diffusion decoder, enabling it to first reason about the structure, meaning, and relationships within a prompt before generating the image itself. This approach allows GLM-Image to excel in scenarios that require precise semantic control, such as generating infographics, presentation slides, posters, and diagrams with accurate embedded text and complex layouts. With a total of around 16 billion parameters, the model achieves strong performance in rendering readable, correctly placed text within images, an area where many image models struggle, while maintaining detailed visual quality and consistency.
Learn more
Inception Labs
Inception Labs is pioneering the next generation of AI with diffusion-based large language models (dLLMs), a breakthrough in AI that offers 10x faster performance and 5-10x lower cost than traditional autoregressive models. Inspired by the success of diffusion models in image and video generation, Inception’s dLLMs introduce enhanced reasoning, error correction, and multimodal capabilities, allowing for more structured and accurate text generation. With applications spanning enterprise AI, research, and content generation, Inception’s approach sets a new standard for speed, efficiency, and control in AI-driven workflows.
Learn more
Mercury Coder
Mercury, the latest innovation from Inception Labs, is the first commercial-scale diffusion large language model (dLLM), offering a 10x speed increase and significantly lower costs compared to traditional autoregressive models. Built for high-performance reasoning, coding, and structured text generation, Mercury processes over 1000 tokens per second on NVIDIA H100 GPUs, making it one of the fastest LLMs available. Unlike conventional models that generate text one token at a time, Mercury refines responses using a coarse-to-fine diffusion approach, improving accuracy and reducing hallucinations. With Mercury Coder, a specialized coding model, developers can experience cutting-edge AI-driven code generation with superior speed and efficiency.
Learn more
ByteDance Seed
Seed Diffusion Preview is a large-scale, code-focused language model that uses discrete-state diffusion to generate code non-sequentially, achieving dramatically faster inference without sacrificing quality by decoupling generation from the token-by-token bottleneck of autoregressive models. It combines a two-stage curriculum, mask-based corruption followed by edit-based augmentation, to robustly train a standard dense Transformer, striking a balance between speed and accuracy and avoiding shortcuts like carry-over unmasking to preserve principled density estimation. The model delivers an inference speed of 2,146 tokens/sec on H20 GPUs, outperforming contemporary diffusion baselines while matching or exceeding their accuracy on standard code benchmarks, including editing tasks, thereby establishing a new speed-quality Pareto frontier and demonstrating discrete diffusion’s practical viability for real-world code generation.
Learn more
