Persistent files and memory
Your conversations, MEMORY.md, USER.md, files, skills, and scheduled jobs survive restarts, upgrades, and renewals.
A private, always-on AI agent with its own workspace, memory, files, scheduled jobs, and model access included. Open the native Hermes workspace in your browser. No servers to run.
Not a chat window with amnesia. A full Hermes install with tools, skills, memory, a real filesystem, and a browser workspace — provisioned for you and kept online.
Your conversations, MEMORY.md, USER.md, files, skills, and scheduled jobs survive restarts, upgrades, and renewals.
Open the native Hermes web workspace and admin dashboard. Configure models, tools, MCP servers, credentials, and channels just like a local install.
Your own filesystem, network, process space, and inference identity. No shared profiles. No other customer can reach your workspace.
Connect your own bot credentials and keep the gateway online in the cloud. Your phone stays a message away from your agent.
Cron jobs, standing instructions, and background work keep ticking even when your laptop is closed.
Restart your runtime, download a full data export, update your password, and manage your subscription from one portal.
Provisioning is automatic. We create a private instance, attach storage, issue an instance-scoped model credential, and hand you the keys.
Pick your login, complete Stripe checkout, and keep the card on file for renewals. Cancel anytime.
A dedicated container gets its own data volume, isolated bridge network, login token, and a model credential that only works for your instance.
We verify the native web workspace and a streaming tool call before marking it ready. Then we email your portal link.
Chat in the browser, connect Telegram or Discord, install skills, schedule jobs, and keep your data across renewals.
No setup fee, no per-seat games. Your subscription keeps the same instance and data alive month after month.
Your own Hermes instance, online 24/7, with model access included.
Secure checkout by Stripe. Cancel anytime.
Yes. Each customer gets a separate Docker container on a separate bridge network with its own data directory. There is no shared Hermes profile, database, workspace, or inference credential. Customer containers cannot open each other’s workspace.
In your instance’s persistent data volume. Restarts and container upgrades keep the same volume. You can download a full export from the customer portal at any time.
Yes. Open the Hermes admin dashboard and add your own bot credentials under Channels. Your gateway stays online in the cloud. You can also connect other Hermes-supported messaging platforms.
Your instance is preconfigured with the TOKNADO OpenAI-compatible endpoint. The default model is toknado/auto. You can change model settings inside your instance, but model traffic is routed through your instance-scoped credential and the hosted relay.
You keep access through the end of the paid period. After that the instance is suspended, but your data is retained for an export window. You can export from the portal before it ends.
Stripe retries the charge. We keep the instance online during a 3-day grace period. If payment still fails, the gateway is stopped and inference is revoked until billing is resolved.
Yes. The native Hermes admin surface can run commands and configure the environment inside your own container. That is why each customer gets a dedicated container instead of a shared profile — the container is the security boundary.
Persistent storage survives restarts, and the host performs routine snapshots. You can also export your full data directory at any time. This is managed hosting, not a regulated backup service; keep your own copy of anything critical.
Launch a private Hermes instance in minutes. Keep it for as long as it earns its keep.