SuperAGI

1.1 Usage

Method 1: config.yaml (Recommended)

When deploying SuperAGI, edit config.yaml:

# config.yaml

# OpenAI-compatible API base URL
OPENAI_API_BASE: "https://api.hpc-ai.com/inference/v1"

# API Key
OPENAI_API_KEY: "sk-your-hpc-ai-api-key"

# Default model name
OPENAI_MODEL: "minimax/minimax-m2.5"

Restart SuperAGI:

docker compose up -d

Method 2: GUI Configuration

1. Start SuperAGI and open the Dashboard (default http://localhost:3000)

2. Go to Settings → Models → Custom Model

3. Fill in:

   • API Base URL: https://api.hpc-ai.com/inference/v1

   • API Key: sk-your-hpc-ai-api-key

   • Model Name: minimax/minimax-m2.5

4. Save

Method 3: Environment Variables

Set in docker-compose.yaml or .env:

# environment section in docker-compose.yaml
services:
  superagi-backend:
    environment:
      - OPENAI_API_BASE=https://api.hpc-ai.com/inference/v1
      - OPENAI_API_KEY=sk-your-hpc-ai-api-key
      - OPENAI_MODEL=minimax/minimax-m2.5

1.2 Creating Agents

After configuration, when creating an Agent:

1. Go to Agents tab

2. Click Create Agent

3. Select the configured model under LLM Model

4. Set Agent goal, tools, etc.

5. Start Agent

1.3 Verify Configuration

# Check SuperAGI backend logs for API connectivity
docker compose logs superagi-backend --tail 50

1.4 References

• Local model config issue: https://github.com/TransformerOptimus/SuperAGI/issues/411

• SuperAGI Documentation: https://superagi.com/docs/Installation/