# OpenAI

Canonical: https://mudpie.ai/authors/openai/
Updated: 2026-09-20

OpenAI publishes the primary product and developer documentation for ChatGPT, Codex, Site Tools, workspace agents, the Agents API and ChatGPT advertising.

Source: https://openai.com/

Description is limited to products documented on official OpenAI pages; image comes from the official GitHub organization. Checked September 19, 2026.

## Referenced in 101

- [WebMCP in OpenAI developer documentation](https://developers.openai.com/codex/webmcp) — ChatGPT Site Tools expose a subset of WebMCP, so support must be checked against OpenAI's documented surface rather than assumed from the draft.
- [Using Site Tools in the ChatGPT desktop app](https://help.openai.com/en/articles/20001423-using-site-tools-in-the-chatgpt-desktop-app) — Site Tools depend on the current page, account and model, and the user remains in control when actions need review.
- [Introducing ChatGPT agent](https://openai.com/index/introducing-chatgpt-agent/) — ChatGPT agent combines research and action, moving the assistant from answering questions toward completing multi-step tasks with user oversight.
- [Agentic Commerce Protocol](https://openai.com/index/introducing-agentic-commerce-protocol/) — ACP gives merchants and agents a shared transaction contract so discovery can become checkout without handing the entire customer relationship to one marketplace.
- [Buy it in ChatGPT](https://openai.com/index/buy-it-in-chatgpt/) — ChatGPT shopping turns product answers into a purchase surface, making feed quality, merchant terms and checkout compatibility part of agent distribution.
- [Memory and new controls for ChatGPT](https://openai.com/index/memory-and-new-controls-for-chatgpt/) — Useful personalization requires persistent context plus controls to inspect, remove, disable and separate what the assistant remembers.
- [Scheduled tasks in ChatGPT](https://help.openai.com/en/articles/10291617-scheduled-tasks-in-chatgpt) — A personal assistant becomes proactive when work can run later, but recurring execution needs visible schedules, notifications and easy cancellation.
- [Introducing Operator](https://openai.com/index/introducing-operator/) — Computer use lets an assistant act through existing websites, but confirmations, takeover and task boundaries remain central to safe completion.
- [Reimagining advertising with AI](https://openai.com/index/reimagining-advertising-with-ai/) — OpenAI frames ads as a financing and discovery layer that must remain clearly labeled and separate from the model's answer.
- [Ads in ChatGPT: the basics](https://help.openai.com/en/articles/20001207-ads-in-chatgpt-the-basics) — The basics page is the product contract: eligibility, placements, labeling, controls and the conditions under which an ad can appear.
- [Sponsored agents in ChatGPT ads](https://help.openai.com/en/articles/20001524-sponsored-agents-in-chatgpt-ads) — A sponsored agent is an interactive commercial experience, not a conventional display unit, so the useful action and disclosure both matter.
- [New ways to buy ChatGPT ads](https://openai.com/index/new-ways-to-buy-chatgpt-ads/) — OpenAI is expanding buying models and formats as the ad product matures, which makes current unit and optimization assumptions explicitly temporary.
- [ChatGPT Ads Manager availability](https://help.openai.com/en/articles/20001245-ads-manager-availability) — Availability varies by market, account and rollout stage, so access should be verified before a media plan assumes the surface exists.
- [Create campaigns for ChatGPT](https://help.openai.com/en/articles/20001210-create-campaigns-for-chatgpt) — Campaign setup establishes objective, audience, budget and delivery before creative; those choices determine what the system can optimize.
- [Create ads for ChatGPT ads](https://help.openai.com/en/articles/20001212-create-ads-for-chatgpt-ads) — Creative must make one useful commercial next step clear while meeting format, landing-page and policy requirements.
- [Write context hints for ChatGPT ads](https://help.openai.com/en/articles/20001521-write-context-hints-for-chatgpt-ads) — Context hints describe the situations where an offer is useful; they are not a hidden keyword list or permission to influence the answer.
- [Conversion-optimized campaigns](https://help.openai.com/en/articles/20001412-conversion-optimized-campaigns) — Conversion optimization only becomes meaningful when the event is reliable, sufficiently frequent and connected to the business outcome the campaign should pursue.
- [Measure results in ChatGPT Ads Manager](https://help.openai.com/en/articles/20001214-measure-results) — Ads Manager reports delivery and attributed outcomes under its own definitions; totals are not automatically interchangeable with analytics or backend revenue.
- [Conversion measurement](https://help.openai.com/en/articles/20001409-conversion-measurement) — A useful conversion setup has a stable event, deduplication, consent and a window that matches the buying cycle.
- [Advertiser guidance for OpenAI web crawlers](https://help.openai.com/en/articles/20001243-advertiser-guidance-for-allowing-openai-web-crawlers) — Ad review and destination understanding can depend on crawler access; blocking the wrong bot can break delivery without saying anything about organic inclusion.
- [Troubleshoot ChatGPT ads](https://help.openai.com/en/articles/20001217) — Diagnose approval, delivery, tracking and destination problems separately rather than treating every zero as a targeting failure.
- [Introducing the Agents API](https://openai.com/index/introducing-the-agents-api/) — OpenAI's Agents API exposes managed cloud agents powered by the Codex harness, with long-running sessions, environments and orchestration behind one API.
- [Introducing workspace agents in ChatGPT](https://openai.com/index/introducing-workspace-agents-in-chatgpt/) — Workspace agents package repeatable, connected workflows for teams and run them in the cloud through ChatGPT's shared work surface.
- [Workspace agents](https://openai.com/academy/workspace-agents/) — A managed team agent combines instructions, knowledge, tools, publishing controls and repeatable runs inside a shared workspace.

This is a Mudpie reference profile, not an account or endorsement. The profile records an authority connected to the source; it is not necessarily the article byline.
