Skip to main content
Back to Glossary
Infrastructure

API (Application Programming Interface)

A way for software programs to communicate with each other, allowing developers to integrate AI capabilities into their applications.


From the archive. This restored glossary entry is educational, not a current vendor comparison. Product names, models, capabilities, pricing, and availability may have changed; verify operational details with primary sources.

APIs for AI

When you use ChatGPT's website, you're using their interface. When you build an app that uses GPT-4, you use their API.

APIs let developers:

  • Send prompts and receive responses programmatically
  • Build custom interfaces and workflows
  • Integrate AI into existing applications
  • Automate AI-powered tasks

Common AI APIs

OpenAI API: Access to GPT-4, DALL-E, Whisper, and more.

Anthropic API: Access to Claude models.

Google AI APIs: Gemini, PaLM, and various specialized models.

Hugging Face: Thousands of open-source models via unified API.

API Pricing Models

Most AI APIs charge per usage:

  • Per token: Pay for input and output tokens
  • Per request: Fixed cost per API call
  • Per minute/hour: Time-based for certain services
  • Subscription: Fixed monthly cost for quota

Understanding pricing matters for production use where costs can scale quickly.

API vs Interface

Use the interface when:

  • Exploring capabilities
  • Occasional personal use
  • Tasks don't require integration

Use the API when:

  • Building applications
  • Automating workflows
  • Need custom behavior
  • Volume requires better pricing

Related Terms

More in Infrastructure