A Chinese AI startup claims it has developed a wearable device capable of translating pets’ vocalizations and behaviors into human language with up to 95 percent accuracy, sparking widespread online skepticism as the internet debates its feasibility.

The company, PettiChat, describes itself as the “world’s first real-time pet translator” and markets a small wearable collar that attaches to a pet’s harness and syncs with a smartphone app. According to its press materials, the AI model was trained on over one million vocal and behavioral samples across multiple animal breeds, claiming 94.6 percent real-time translation accuracy and automatic conversion in under 1.2 seconds.

The startup asserts that its device enables two-way communication between pets and owners, translating pet sounds into human language while also converting spoken words into sounds pets instinctively recognize. It further states the product has undergone testing on more than one thousand dogs and cats over a two-year period and currently raises funds via Kickstarter.

Critics immediately questioned the claim’s practicality. The internet widely dismissed the 95 percent accuracy figure as marketing rather than verified science, noting that simple pet behaviors—like barking at the front door or vocalizing at 3 a.m.—do not require complex AI translation to understand. Industry analysts emphasized that while decoding basic emotions through sound and behavior is plausible, translating specific English sentences from animal vocalizations remains scientifically unproven.

PettiChat’s materials highlight the device as a bond-building tool with chat history features in its companion app, but skeptics warn early adopters may receive only an entertaining novelty rather than a true breakthrough in human-animal communication. Until independent testing confirms the technology’s real-world utility, the claim remains firmly in the realm of viral marketing.