MultiverseAI Logo
    Developer API

    Build with the MultiverseAI API.

    One OpenAI-compatible API for every model in MultiverseAI — plus your agents, workflows, tasks and the full agent toolkit.

    [ 97 agent tools · 1 credit = $0.01 ]
    Quick start

    A key, a base URL, and you're calling it.

    Create a key in Developer Settings, then call the API with any HTTP client or the official openai SDK.

    Example request
    curl https://ralcxfiefghnpadaddzu.supabase.co/functions/v1/public-api/v1/chat/completions \
      -H "Authorization: Bearer $MULTIVERSE_API_KEY" \
      -H "Content-Type: application/json" \
      -d '{
        "model": "openrouter/anthropic/claude-opus-5.5",
        "messages": [
          { "role": "user", "content": "Explain quantum computing in two sentences." }
        ]
      }'
    Endpoints

    A few of the endpoints.

    The docs cover all of them — these are the ones most integrations start with.

    1. [01]
      POST/v1/chat/completions

      OpenAI-compatible chat across every model, with streaming, image input and agent tools

      modelmessagesstreamtoolstemperaturemax_tokens
    2. [02]
      POST/v1/images/generations

      Generate images with any image model, from a prompt or reference images

      promptmodelnimage
    3. [03]
      POST/v1/videos

      Start a video generation job, then poll GET /v1/videos/{id}

      promptmodeldurationaspect_ratioimage_url
    4. [04]
      POST/v1/workflows/{id}/run

      Run one of your saved workflows

      inputswait
    5. [05]
      POST/v1/agents/{id}/chat

      Chat with one of your AI agents, with saved conversation history

      messageconversation_idstream
    6. [06]
      POST/v1/queue

      Queue a prompt or workflow to run now or at a scheduled time

      promptworkflow_idpriorityscheduled_for
    7. [07]
      POST/v1/tools/{name}

      Run a MultiverseAI agent tool directly

      argumentsconfirm
    8. [08]
      GET/v1/models

      List every available chat, image and video model

      modality
    What you get

    Everything in your account, from code.

    The API is the same surface the app and the macOS client use — nothing is held back for a higher tier.

    A blurred corridor of white server racks with small status lights
    [01]

    OpenAI-compatible

    Use the official openai client or any compatible library — just change the base URL and key

    [02]

    Every model in one place

    The same chat, image and video models as the MultiverseAI app

    [03]

    Agent tools built in

    Turn on tools: true and the model can search, scrape, run workflows and more

    [04]

    Streaming

    Server-sent events for chat, agents and conversations

    [05]

    Scoped keys

    Switch workflows, agents and tasks access on or off per API key

    [06]

    Pay with credits

    Usage is paid from your MultiverseAI credits — check your balance at /v1/credits

    Libraries and tools

    Plain HTTP today. SDKs on the way.

    Use any OpenAI-compatible client now; the TypeScript SDK and CLI are not on npm yet.

    REST / HTTP

    Any language with an HTTP client

    [ available ]

    OpenAI-compatible clients

    Official openai SDKs and compatible libraries

    [ available ]

    TypeScript SDK

    @multiverseai/sdk — not on npm yet

    [ coming soon ]

    CLI

    multiverse-cli — not on npm yet

    [ coming soon ]
    API pricing

    Pay for what you use. Nothing you don't.

    Usage is billed at provider cost + 50%; fractions carry over until they add up to a whole credit.

    Pay As You Go
    $0.01/credit

    Perfect for getting started

    • All AI models: provider cost + 50%
    • 100 free credits to start
    • Community support
    • Developer keys with per-area permissions
    Start Building

    Ready to build with AI?

    Start building today with 100 free credits; packs from $10 for 1,000 credits.

    No credit card required
    100 free credits
    Same credits as the app