View Categories

Before You Begin

Requirements #

  • WordPress 6.0 or newer.
  • PHP 8.0 or newer.
  • Gravity Forms installed and active.
  • A paid GravityOps AI Feed plan that supports the feature you want to use.
  • An Anthropic API account for text generation.
  • An OpenAI API account for image generation.
  • Gravity Flow only if you want to use AI workflow steps.

Provider Accounts Are Separate #

A Claude web subscription does not automatically supply Anthropic API credits. Likewise, a ChatGPT subscription does not automatically supply OpenAI API credit. Create and fund the provider’s API account, then create an API key in that provider’s developer console.

Generation stops if the provider account has no available credit, the key lacks permission, or the provider enforces a usage or rate limit. Use the Test Anthropic and Test OpenAI buttons after saving keys, but remember that a successful connection test does not guarantee sufficient credit for every model or request.

Prepare The Form First #

Before creating a feed, add the fields that will supply prompt values and receive generated results.

For text generation, add a Text, Paragraph, Hidden, or other suitable destination field. For multi-field output, create every destination field before building the mapping.

For image generation, add a File Upload field. A post-submission image feed cannot target a required File Upload field because Gravity Forms validates required fields before the AI feed runs. Leave that target field optional.

Data And Cost Review #

Any value inserted into a prompt with a merge tag is sent to the selected provider. Review the form for personal, financial, health, legal, or otherwise sensitive information before enabling generation. Tell users when their information will be processed by an external AI provider when applicable to your site and jurisdiction.

Provider charges depend on the selected model, prompt size, output length, image size, and image quality. Start with a test form and modest limits. Monthly text token caps and public frontend request limits can reduce accidental usage, but they are not a substitute for provider-side budgets and alerts.