Skip to main content

Documentation Index

Fetch the complete documentation index at: https://agent37.com/docs/llms.txt

Use this file to discover all available pages before exploring further.

You create an agent with one call, then talk to it with the next. Files, connected accounts, and memory stay on the agent until you delete it.
The Agents API is in a private beta with a small group of design partners. Book a call to see if it’s a fit and we’ll get you set up with access.

​
Prerequisites

  • An API key from the private beta (see above). Send it as a bearer token. Keep it server-side. Never ship it to a browser or end user.
  • A funded balance. Your starter credit covers your first agent; once you are onboarded you can top up in the dashboard for more.
1

Create an agent

POST /v1/instances returns an instance that is ready to take a message the moment its status is running. Every field is optional, so a POST with no body works.
curl -X POST https://api.agent37.com/v1/instances \
  -H "Authorization: Bearer sk_live_..."
# -> { "id": "inst_x7", "status": "running", ... }
2

Send it a message

Send input, passing the instance’s id as instance_id. Set stream: true to receive Server-Sent Events as the agent reasons, calls tools, and writes its answer.
curl -N https://api.agent37.com/v1/responses \
  -H "Authorization: Bearer sk_live_..." \
  -H "Content-Type: application/json" \
  -d '{
    "instance_id": "inst_x7",
    "input": "Research the top 3 EV makers, write a memo.",
    "stream": true
  }'
3

Continue the conversation

The first reply returns a session_id. Reuse it to continue the same thread: the agent keeps the full history, so you send only the new input.
POST /v1/responses
{ "session_id": "sess_1", "input": "Make it shorter, add a quote." }
Prefer not to stream? Send stream: false (the default) and the call returns the finished response as one JSON body.

​
Next steps

Core concepts

Instances, sessions, responses, and the agent-vs-model distinction.

Send a message

Every field on the core call, plus the response shape.

Streaming events

The full event list and a client parser.

Instances

Size, manage, and delete the agent’s computer.