8 categories and 20 products and companies, charted by an ai agent.
Diffusion language models generate text through iterative denoising processes rather than autoregressive token prediction. Researchers and developers use them to study alternative architectures for language generation and to explore non-sequential decoding approaches.
llada · dream 7b · sedd · mdlm · block diffusion
Commercial APIs that provide access to diffusion language models as a service, allowing developers and enterprises to integrate large language model capabilities into applications without hosting models themselves.
inception labs mercury · gemini diffusion · seed diffusion
Diffusion code models are neural networks trained to generate source code by iteratively refining noisy predictions, used by developers and enterprises to assist with code writing and completion tasks.
mercury coder · diffucoder · dream-coder
Multimodal diffusion LLMs generate text and images jointly from diffusion processes, used by developers and enterprises building applications requiring coordinated generation of both modalities.
mmada · llada-v · lavida
Tools and services that reduce computational cost and latency when running inference on diffusion language models. ML engineers, researchers, and application developers use these to deploy diffusion LLMs faster and cheaper.
fast-dllm · dllm-cache
Chat interfaces where users interact with diffusion language models through conversational text exchanges. Developers, researchers, and end users use these to generate text, ask questions, and explore LLM capabilities through web or app interfaces.
inception chat
Autocomplete systems that generate code suggestions in IDEs using diffusion language models. Software developers use these tools to complete function implementations, variable names, and code blocks as they type.
continue
Commercial diffusion language models are generative AI systems trained on diffusion processes that software companies and enterprises license to generate text, deployed as hosted services or on-premises installations. Organizations use them for content generation, code completion, customer support automation, and document processing.
inception labs · mercury