Hands-on Operator test
Casey Newton tested OpenAI's Operator on walking tours, a Gatsby lesson plan, Instacart groceries, and an Uber price check.
Operator handled tour search and lesson plan but was slow; adding three grocery items took 15 minutes and an Uber price check took five.
Overview
How it was set up
OpenAI Operator, a research preview in the $200-a-month ChatGPT Pro tier, running in its own browser with screenshots and mouse actions.
Worth knowing
It uses its own browser, so it lacks logins and preferences, and some sites block it. Free ChatGPT did the simple tasks faster.
A prompt to start from
The source does not show the prompt.
Sources 1
History
- Source published
- 2025-01-24
- Added to Agentic Feats
- 2026-10-13
- Pick
- No
- Markdown copy
- /c/hands-on-operator-test.md
More in research
36 featsResultIt created the text file as asked, but after clicking a result it took a screenshot too early and hallucinated the answer.
ResultThe draft was usable but weaker than accepted NSF summaries; thinking mode helped, though it began hallucinating and misreading the sources.
ResultOperator collected data for about 25 of 100 schools before the author stopped it; o1 then built a working chart dashboard from that data.