Skip to content

AI tools

Hermes Agent

Free softwareLocal

Run a personal agent for messaging, files, memory and scheduled tasks on your own server.

What you can use it for

Run a personal agent for messaging, files, memory and scheduled tasks on your own server.

  • Personal agent accessible through messaging apps.
  • Local-model endpoint support and configurable tools.
Workflow coverage and comparison notes

Self-hosted persistent personal-agent workflow: messages, file organisation and configurable scheduled tasks. Hosted services and advanced integrations are separate. Also covers the self-hosted personal-agent slice of OpenAI Dot; not its managed cloud computer or complete app integrations.

Before you switch

Setup and maintenance are required. Local inference needs capable hardware and at least 64K model context; task quality depends on the model. Hosted models, browser/search services and servers may cost extra. A firsthand user reports ongoing debugging. No managed cloud computer or complete Dot/Grok integration parity. Replace the solo persistent messaging-assistant slice of Bolter with a self-hosted agent for files, memory and configured routines. Free MIT software; local tool-capable models require suitable owned hardware and electricity. Server hosting, cloud model APIs and optional external tools can cost extra. Not Bolter’s managed multiplayer chat, external-agent federation, no-code blueprint network or equivalent approvals/receipts; connector and model quality vary.

Alternatives to these paid apps

Each comparison covers the workflow described above, rather than every paid feature.