liveBook · Manning

OpenAI API

As explained in Vibe Engineering

The OpenAI API is a public programming interface for accessing OpenAI-hosted artificial-intelligence models and services.

The OpenAI API connects an application to hosted language models, but integration is an engineering tradeoff: stronger models may improve accuracy while increasing token cost. Keep the provider replaceable so experiments can compare OpenAI with other hosted or self-hosted models. section 6.2

A disciplined integration makes the request and its failure modes explicit:

  • Configure the API key, model, and timeout through YAML or environment variables; running the application requires account credits. section 6.2

How Vibe Engineering covers it

How deep it goes
Explains configuration, prompt assembly, chat-completion parameters, response handling, and basic error paths.
Examples
Uses a Java SQL generator with YAML configuration, schema-aware prompts, chat completions, and markdown-stripping cleanup.
What it assumes
Assumes an OpenAI account with credits and an application capable of supplying configuration, prompts, and service calls.
What you can do after
Operate/configure an OpenAI-backed SQL generator with selectable models, bounded responses, prompt context, and basic failure handling.
How it connects
Connects LLM providers, prompt construction, SQL generation, environment variables, and model-cost tradeoffs.

Also explained in

Compare how 6 books explain OpenAI API →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Vibe Engineering. The section links point to the parts of the book it draws on.