A small model that works like a big one. Memory that never overflows, tools that ask first, and a personality that's yours.
Every turn is built fresh from a memory that never forgets. The context stays a scratchpad.
Searches the web, reads each source whole, and tells you which one said what.
Hand Axom a goal and walk away, or have it keep watch on something for you.
Tone, length, humor, format. Set them yourself or let Axom learn how you write.
Reads in your project are free. Writes and commands ask first, per project.
No GPU? Run the model on a node while your memory and files stay home.