What OpenAI announced
OpenAI introduced GPT-6.1 Terra on 26 October 2026, describing a model for visual reasoning, precise computer use and clear writing. Its documentation lists availability in ChatGPT Work and Codex, subject to plans and workspace controls.
This matters for tasks involving screen interpretation and interaction. A model announcement and a feature's availability in an application are distinct: each product determines the tools and access it provides.
A useful test for a catalogue team
As an iabot editorial proposal, we would compare ten product photos against their listings and prepare possible discrepancies. The owner would supply images, correct attributes and review criteria. The result would list the reference, observation and visible evidence.
A second stage could draft corrections for review. Before allowing store changes, check that the employee distinguishes a suggestion from an authorised edit. The team can review more material while a person retains commercial judgement and publication decisions.
How to decide whether it fits
Measure correct findings, false alarms and human validation time. Include difficult images, incomplete data and similar screen elements. A correct screenshot is insufficient if the final record is saved incorrectly.
Use the same inputs and objective when comparing alternatives. This article describes a launch and an evaluation proposal, not a Terra test in Hub Bot. iabot aims to expand team capability through clear tasks and explicit permissions.
Sources and documentation
Provider announcements and results are attributed to their sources. Use cases are editorial proposals, not our own tests or confirmation of model availability in iabot.