What is OpenAI?

What it is

OpenAI builds large language models, which are programs trained on large amounts of text to write, summarize, translate and answer questions. The main product is ChatGPT, a chat app where a person types a request and reads the reply. The same models are also sold to developers through an API, which lets a company's own software send text in and get text back.

The API is separate from the chat app: a developer connects it to your systems, and it has its own account, keys and bill. OpenAI also makes a coding agent called Codex. All of these tools can produce fluent text that is wrong.

Why it matters to a business

• Staff already use it: people try a chat app on their own, so write a rule on what may be pasted into it.
• Data handling: text sent to any AI service leaves your network, so sensitive records need a decision before they are shared.
• Cost per use: the API is billed by usage, so a runaway script can run up a large charge.
• Review: someone accountable should read AI output before it is acted on or sent to a customer.

How KawaConnect works with OpenAI

Our AI and automation work includes tools that call AI models through their APIs, and OpenAI models are one option among several. We choose the model for each task. Our engineers use Codex in software work, and a person reviews every change before it ships. We keep each application's key separate and stored safely, as described in AI API keys. We are not an OpenAI partner or reseller.

Learn more

• OpenAI on WikipediaA neutral overview of the company and its products, useful for readers who want background.
• OpenAIThe company's own site, with its current products.
• OpenAI API documentationThe developer reference, useful for the person connecting the API to other software.