Spanish Firm's New Model Faces Criticism for Undisclosed Chinese Origin

On September 18, Multiverse Computing, a Spanish AI company, announced on X the release of its new large language model, Quasar 1.1 438B, describing it as "the first AI model trained on data generated by quantum computing." The company said the upgraded version of what it had previously called "Europe's finest model" is more compact and efficient.

Soon after the announcement, users questioned the European origin claim. One user pointed out that the company's technical documentation states that Quasar 438B was built on GLM-5.2, an open-weight model developed by Chinese AI firm Z.ai (Zhipu AI), and then compressed using its own pipeline. This prompted criticism that labeling it a "European model" was misleading.

The claim was confirmed in the company's own article about Quasar 1.1 438B, where it acknowledged that Quasar 438B is based on GLM-5.2 from Z.ai. Many commenters mocked the company for not disclosing the model's Chinese origins and advised it to credit Z.ai out of respect.

Earlier on September 2, when Multiverse Computing released Quasar 438B, it touted the model as setting a new benchmark for European AI and awarded it a high score on its "AI Analysis Index," but made no mention of its connection to GLM-5.2. At that time, some users on X and Hacker News accused the company of misleading marketing.