POST
/
agent
curl --request POST \
  --url https://api.openmind.org/api/core/agent \
  --header 'Content-Type: application/json' \
  --header 'x-api-key: <api-key>' \
  --data '{
  "model": "gpt-4.1-nano",
  "message": "Analyze the market trends for renewable energy",
  "response_model": {}
}'
{
  "content": {},
  "usage": {
    "prompt_tokens": 123,
    "completion_tokens": 123,
    "total_tokens": 123
  },
  "model": "<string>"
}

​
Multi-Agent LLM Integration

The Multi Agent plugin seamlessly integrates OpenMind’s OM1 capabilities within an agent architecture, allowing you to leverage our agent-based architecture using multiple LLMs for a rich experience.

​
Overview

Multi Agent routes requests through OpenMind’s collaborative multi-agent system, providing a powerful way to solve complex tasks. This approach combines the flexibility of multiple specialized agents working together to deliver enhanced capabilities.

​
Key Features

  • Collaborative Intelligence: Multiple agents work together to solve complex problems
  • Robotic Team Integration: Harness the power of specialized agents working together
  • Compatible Interface: Uses the same patterns as standard LLM implementations

​
Getting Started

To run the multi-agent system on OM1:

uv run src/run.py multiagent

​
Technical Details

Under the hood, MultiLLM sends requests to the agent endpoint at https://api.openmind.org/api/core/agent. The system formats your inputs appropriately, facilitating communication between specialized agents like navigation and perception.

The request includes:

  • System prompt context
  • User inputs
  • Available actions for the agents

​
Use Cases

  • Robotics Applications: Interface with physical systems through natural language
  • Complex Decision Making: Combine multiple specialized agents for better reasoning
  • Collaborative Problem Solving: Leverage different agents in a single workflow.

Authorizations

​
x-api-key
string
header
required

Body

application/json

Response

200
application/json

Successful agent run response

The response is of type object.