Chat
Hermon AI
The primary Hermon experience for writing, learning, reasoning, coding, and getting work done.
Start a chat →The Hermon ecosystem
Everything around the conversation.
Hermon begins with a large language model and grows into persistent projects, agents, developer APIs, local runtimes, and an open reference engine. Chat stays at the center.
Workspace
Projects and agents
Organize persistent work, reusable agents, capability cells, and environments around every conversation.
Open workspace →Developers
API and open source
Build on Hermon through an OpenAI-compatible API, the application platform, and the reference inference engine.
Explore GitHub →Application platform
Conversations become durable work.
Hermon Web adds identity, chat history, projects, agents, tools, and policy around the model without putting infrastructure in the user's way.
Application platform repository →Inference
A runtime path Hermon can own.
The Hermon reference engine develops scheduling, paged memory, deterministic attention, native kernels, and device execution behind measured correctness gates.
Read the engine architecture →Explore more