Consume AI models through monday

You can now consume leading AI models directly through monday.com — no separate vendor accounts, API keys, or procurement cycles. Authenticate with your monday API token, call familiar endpoints, and let monday handle routing to upstream providers (OpenAI and Anthropic) while metering usage against your workspace.

There are two ways to consume models, depending on how much control you need.

Two ways to consume models

Both run on the same monday AI gateway, so they share the same access requirements and token consumption. Pick run_prompt when you just need a prompt-in, text-out call inside an existing GraphQL request; pick the Models API when you need streaming, conversation state, tools, or non-text modalities.

Models API (OpenAI-compatible REST)

The Models API matches the OpenAI HTTP API, so you can reuse existing OpenAI tutorials and SDKs by swapping only the base URL and API key.

curl -sS -X POST "https://api.monday.com/platform-ai-gateway/openai/v1/chat/completions" \
  -H "Authorization: Bearer <MONDAY_API_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "monday-standard",
    "messages": [{ "role": "user", "content": "Reply with the word: ok" }]
  }'

It covers four capabilities through one API: POST /chat/completions, POST /embeddings, POST /images/generations, and POST /audio/speech.

run_prompt (GraphQL)

A simplified, GraphQL-native entry point to the same gateway. Send a prompt and get the generated text back in the same GraphQL request — no separate base URL or OpenAI client required. See the AI reference.

mutation {
  run_prompt(
    prompt: "Reply with the word: ok"
    config: { model: MONDAY_FAST }
  ) {
    content
  }
}

Using Make.com (no code)

If you build automations in Make.com, you can consume monday AI models without writing any code using the Run monday.com AI (beta) module. It wraps the same models API, so it uses your monday connection for authentication and draws down the same monday AI tokens — giving you the same spend tracking and governance inside your Make scenarios.

App Developers

This is also available for apps built on the app framework — add the AI:Consume scope and authenticate with the app token to consume models on behalf of the installing account.

Further reading

• Models API overview — positioning, metering, and feature summary.
• Getting started with the Models API — token, base URL, capabilities, and first request.
• Models API supported models — model IDs, aliases, and defaults.
• run_prompt (GraphQL) — single-turn chat completions in the core API.