OpenAI seems to be testing out another feature called “Branch from here” inside the ChatGPT web app. The feature was spotted by user Tibor Blaho and shared on X (formerly Twitter), showing what looks like an early version of a branching tool for conversations.
In the screenshot Blaho posted, the feature appears in the response menu, sitting next to the usual options like thumbs up, thumbs down, copy, and share. The “Branch from here” button seems designed to allow users to split off a response into a new thread, preserving the original while exploring a different direction. Blaho noted that the interface also hints at labels such as “Branched from {conversationTitle}”, indicating OpenAI plans to make it clear where the new chat originated.
The community quickly began speculating about how this differs from simply editing a message. One user suggested it might just be a minor UI change. However, Blaho pointed out that the branching feature appears to create entirely new conversation threads compared to just replacing text.
If rolled out broadly, the update could make ChatGPT more useful for people juggling multiple ideas within the same discussion. Right now, starting a new thread means copying and pasting prompts, which isn’t ideal for complex workflows.
However, it’s worth noting that OpenAI often tests several features on its platform. That does not guarantee that the feature will make it into the final app for everyone. However, this feature in particular seems like a good update and will surely benefit the users.
That’s all about OpenAI and ChatGPT for now. While the company is testing out smaller updates, it has also started working on the next big GPT-6 model. Recently, CEO Sam Altman hinted that GPT-6 may arrive faster than expected. Speaking in San Francisco, Altman said the upcoming model will be more adaptive, learning user preferences and routines to create more personalized experiences.
