Set up ModelRiver in minutes

1. Sign in and create a project

1. Visit modelriver.com and create an account.
2. Name your first project (for example, “Production” or “Staging”). Projects keep credentials, workflows, and logs isolated.
3. Invite teammates from the user menu if they need dashboard access.

2. Connect AI providers

1. Navigate to Settings → Providers.
2. Paste the API key for each vendor you plan to use (OpenAI, Anthropic, Google, Cohere, Mistral, or a custom endpoint).
3. Save. ModelRiver encrypts credentials immediately. You can rotate or delete them at any time. In the project sidebar, the stacked server icon takes you straight to this view.

3. Create API keys for your apps

1. Go to API Keys inside your project.
2. Click Create Key, give it a descriptive name, and copy the generated value. Keys are shown once—store them in your secret manager.
3. Revoke keys at any time; the change takes effect immediately. Find the key icon tile in the sidebar when you’re ready to mint a new credential.

4. (Optional) Define a structured output

1. Open Structured Outputs and click New Structure.
2. Provide a name, optional description, and the JSON schema the model should follow.
3. Attach the structure to a workflow or reference it directly when calling the API.

5. Build a workflow

1. Head to Workflows and choose Create Workflow.
2. Select the primary provider/model, optionally add up to two fallbacks, and attach a structured output if you need reliable JSON.
3. Add cache fields (for example, user.email or messages.0.content) to surface important request data in responses and logs. The connected nodes icon highlights the Workflows area; after saving, use the play button icon to jump into the Playground and verify the configuration.

6. Make your first request

Use your new API key to call the unified endpoint:

curl -X POST https://api.modelriver.com/v1/ai \
  -H "Authorization: Bearer mr_live_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "workflow": "marketing-summary",
    "messages": [
      {"role": "user", "content": "Summarise this week's launch"}
    ]
  }'

Review the response and open Request Logs to inspect the run—prompt, response, token usage, and cached data are all recorded for you.

Next steps

• Continue with the Dashboard tour to master workflows, logs, and analytics.
• Jump straight to Integrating the API for language-specific examples.
• Learn how ModelRiver secures data in the Security guide.