Skip to article
OpenOx
ModelsApps
Sign inSign up

OpenOxAI

Documentation

From the essentials to your first call.

About OpenOxAI

Introduction to OpenOxAICapabilities and protocolsAPI and StudioWhere to start

Getting started

Account and fundingCreate an API keyIntegration guides
DocsAbout OpenOxAI

Where to start

Choose a reading path for your goal, from checking prices to completing your first request.

On this page
Start with your goalRecommended reading orderBefore you connectHow to verify the integrationWhere to find help

Start with your goal

I want to integrate my applicationReview protocols and addresses, then create a key, select a model and make a request.
I want to configure a clientPrepare your account and key, then choose a guide for your client and operating system.
I want to try models in the browserReview Studio workflows and costs before opening the workspace.

Recommended reading order

  1. Understand pricing

    Check plan availability and usage billing before deciding on a budget and workflow.

    Plans · Usage billing
  2. Prepare the account and funding

    Sign in and check funding options and available amounts in Billing & Credits.

    Account and funding guide
  3. Create a key and confirm access

    Create your key, check its product, group, models and expiry, and copy the exact model ID.

    API key guide · Model catalog
  4. Configure and send a first request

    Set the endpoint and credentials for the protocol, then send a short request and confirm it completes.

    Endpoint addresses · Client guides
  5. Review usage before scaling up

    Check the model, time, tokens and cost. Expand usage once the result matches your expectations.

    Open Usage

Before you connect

  • You can sign in to your OpenOxAI account.
  • The account has funds or actual entitlements eligible for the request.
  • You have an API key that is neither expired nor disabled.
  • You know the exact model ID and the protocol and address required by the client.

How to verify the integration

Check that the request returned the expected content. For streaming, also check that the output completed. HTTP 200 or a successful model-list request alone does not prove that a model generation completed.

Then find the record in Usage by time and model and review consumption. For authentication, access, allowance or connection issues, read the error body first and include the request identifier when asking for help.

Where to find help

Key and access issuesCheck key status, model scope and configuration.
Billing and allowance questionsReview usage reconciliation, billing units and common issues.
Contact supportProvide the time and timezone, model ID, request identifier and error.

Read the articles relevant to your task, then explore advanced topics after your first call.

PreviousAPI and StudioNextAccount and funding

On this page

Start with your goalRecommended reading orderBefore you connectHow to verify the integrationWhere to find help
Need a hand?Contact support