Agent 37

Description

Agent 37 is a platform for hosting always-on AI agents. It runs harnesses such as OpenClaw, Hermes and Claude Code around the clock, either deployed in one click from a dashboard or created for your own users with a single API call. Dashboard instances come with a task board, web terminal, file browser, live desktop and scheduled jobs, so there is no VPS or Docker setup to manage.

Agents connect to the tools users already rely on, including Gmail, WhatsApp, Telegram, Discord, Slack and GitHub, with 1,000+ app integrations available in one click through Composio. Users can bring their own API keys or sign in with existing subscriptions like ChatGPT or Claude, and higher tiers add bundled models with usage limits. The Cloud API creates one isolated sandbox per user, each with its own disk and URL, billed per minute, and the whole experience can be white-labelled under your brand.

Agent 37 states it does not train models on your data or collect your conversations, files or terminal sessions, with encryption in transit and at rest and SOC 2 Type I certification (Type II in progress). There is no lock-in: files and keys live inside your container, your workspace is plain downloadable files, and you can cancel anytime. Dashboard instances start from $3.99/mo across Basic, Plus, Pro and Max tiers, and API instances are metered by the minute.

Agent 37's Core Features

• Host OpenClaw, Hermes or Claude Code agents around the clock

• One-click dashboard deploy with task board, web terminal and file browser

• Cloud API that creates one isolated sandbox per user, billed per minute

• 1,000+ integrations via Composio in one click

• Bring your own API keys or sign in with existing subscriptions

• White-label agents under your own brand

• No training on your data, encrypted in transit and at rest, SOC 2 Type I

• No lock-in with downloadable workspace files and cancel anytime

How to use Agent 37?

1. 1

   Choose a path: Deploy from the dashboard for yourself, or use the Cloud API to create agents for your users.

2. 2

   Pick a harness: Select OpenClaw, Hermes, Claude Code or your own custom harness.

3. 3

   Connect tools: Link Gmail, WhatsApp, Slack, GitHub and 1,000+ integrations through Composio.

4. 4

   Add credentials: Bring your own API keys or sign in with subscriptions like ChatGPT or Claude.

5. 5

   Run: Launch the always-on agent and manage it from the dashboard or via the API.

Agent 37's Use Cases

• Always-on personal agent
• Agents for your product's users
• Workflow automation
• Client agent hosting

FAQ from Agent 37

• What is Agent 37?

  Agent 37 is a hosting platform that runs AI agents like OpenClaw, Hermes and Claude Code around the clock, deployable in one click from a dashboard or created per user through an API.

• Can I bring my own API keys or AI subscriptions?

  Yes. On Basic you bring your own keys, which go from your container straight to the model provider, and you can sign in with subscriptions like ChatGPT or Claude. Higher tiers add bundled models with usage limits.

• Can I connect my own WhatsApp, Telegram or Slack?

  Yes. Your agent talks through your own accounts on WhatsApp, Telegram, Discord, Slack and other channels, plus 1,000+ integrations like Gmail, GitHub and Google Calendar via Composio.

• Do my users each get their own agent?

  Yes. The Cloud API creates one isolated sandbox per user, each with its own disk and URL, from $1.99 per agent per month billed per minute, and you can white-label the whole thing.

• Do you train AI models on my data?

  No. Nothing your agent reads, writes or says is used to train a model, and Agent 37 does not collect your conversations, files or terminal sessions.

• Is there any lock-in?

  No. Your files and keys live inside your container, your workspace is plain files you can download anytime, and you can cancel from the dashboard whenever you like.