humanoid.guide Publishes Article on ‘GR-Dexter’

humanoid.guide published an article titled “Welcome, GR-Dexter!” on November 15, 2026.

What can be confirmed this time is that humanoid.guide published an article with “GR-Dexter” in its title; it is not an official announcement by ByteDance or the presentation of a primary source. The starting point should be not to equate publication of the article with a product announcement or a confirmed development plan. Technical specifications, the form of availability and evaluation methods cannot be supplied based only on a news headline. On November 7, we covered an article by the same outlet about a ByteDance Seed robotic hand: humanoid.guide Publishes Article on ByteDance’s “ByteDexter”. The information provided at the time did not confirm support from an official announcement or primary source. “GR-Dexter” in the title of the current article differs from “ByteDexter” in the earlier one. Similarity between the names alone does not provide grounds to infer that they are the same project, that one is a renamed version of the other or that there is a technical lineage between them. The two should be treated as separate reports until their relationship is confirmed by official materials or other sources. An article by 36Kr that we reported on July 15 also covered ByteDance’s entry into the physical AI field, but the summary provided did not give details of a specific official announcement or figures 36Kr Reports ByteDance’s Entry Into the Physical AI Field. Multiple media articles, including this one, indicate that related reports exist, but they do not establish what the company is developing or offering. It is therefore necessary to assess the accumulation of reports separately from the accumulation of officially confirmed products and research plans. Points to watch include whether ByteDance or the research team releases an official explanation, technical materials such as papers, information about the devices or functions involved, evaluation conditions and the stage of use. Without confirmation of these, it is not possible to discuss implementation status or performance beyond what the name suggests. See the original source for details.