Give it a purpose
Name your agent and tell it how you like to work. Choose a supported model and how you’ll pay for usage.
An agent for your ideas, research, and everyday work. Make it yours with skills. Work solo or with your team, in Studio and the chat apps you use. Let Lomus handle the hosting.
We can explore the possibilities, collect useful sources, and turn your thoughts into a clear first draft.
“Help me draft a quick reply.”
Powered by Hermes. Hosted by Lomus. Channel setup varies; the current self-serve wizard offers Studio and Telegram.
Name your agent and tell it how you like to work. Choose a supported model and how you’ll pay for usage.
Start in Studio, or connect through Telegram, Slack, or Discord. Find a conversation that fits your day.
Explore curated skills, add your own, and refine your agent as your needs change.
Explore a few ways to work with an agent. These are scripted examples, with an illustrative interface.
Your instructions.
Your skills. Your way.
An agent combines a model with instructions and tools. Here are three things to consider:
Start simple. Add the capabilities you need, with a workspace that keeps things in view.
Explore a curated catalog, enable skills for your agent, or upload a custom skill for review. See setup requirements before you start.
Explore skillsUse supported OpenAI or Anthropic models. Bring your own key or use managed credits on eligible plans.
See usage, manage your workspace, and configure budget controls. Know where your agents stand.
Inside your workspaceInvite team members to a shared workspace and work with the same agents and skills. Set roles for owners, admins, members, and operators, so each person has the access they need.
Store API keys and credentials encrypted at rest. Share a secret with your workspace, or keep it private for agents acting on your behalf. Saved values aren’t displayed again in the app.
Voice capabilities and image generation are available on eligible plans. These are agent capabilities, not separate phone or voice channels.
Paid plans from $19/month. Choose the space your agents need, with clear options for model usage.
Find your planYour Lomus workspaceHosting, agent capacity, and the features in your plan.
Your model usageBilled by your provider with BYOK, or paid through eligible Lomus AI credits.
Get help with your first agent or your next idea.
Explore agents, workflows, and lessons learned.
Suggest improvements and discuss feature requests.
More ways to connect. More choice in how you build.
Create a Hermes-powered agent, chat with it in Studio, and connect Telegram, Slack, or Discord. Add instructions and skills to shape what it does. WhatsApp, Google Chat, Microsoft Teams, Lark, and OpenClaw are coming soon.
You can create an agent through the app without writing code. Some skills need additional setup or a service API key. The skill details explain those requirements.
A Trial workspace has one test agent and no included AI credits. Use your own supported model key or purchase eligible Lomus credits for model usage. It is not unlimited free AI.
Yes. Supported chat providers are OpenAI and Anthropic. Starter requires your own key; other eligible plans also support Lomus-managed AI credits. BYOK usage is billed by your provider.
The previews illustrate real product concepts using a refined interface. They are scripted demonstrations, not live agents or a promise that every screen looks identical.
Yes. Invite people to your workspace and assign roles to manage who can run agents, create or configure them, invite members, or access billing. Studio conversations remain personal to their creator; collaboration does not mean a shared chat transcript.
The secrets vault encrypts stored values and does not display them again after saving. Choose workspace-shared or private secrets. Private access is tied to your identity; linked Telegram or Discord identities are needed to use private secrets through those channels.
Visit Lomus Community to ask questions and share ideas, or email support@lomus.ai for account and billing help.
Start with one agent. Give it a purpose. Take it from there.
Create your agent Explore with a Trial workspace. Model usage costs are separate.