Will Meta End the China Open Source Model? Muse Core Member Criticizes Kimi, Triggering Backlash in the Comments Section

According to TrendForce's Beating Monitor, Zengyi Qin, a member of Meta's Super Intelligence Lab and a core contributor to Muse Spark, publicly criticized Chinese open-source models. He stated that Meta has an order of magnitude more compute power and better data, predicting that Muse Spark will eventually surpass Chinese models like Kimi.

He also extended his analysis to the business realm: major U.S. clients such as JPMorgan will switch to U.S. open models for compliance, causing Chinese labs to lose this portion of their inference revenue. While Meta can rely on revenue from Facebook and Instagram, Chinese model companies are more dependent on model-related income.

The comments section quickly raised counterpoints: If Meta has not lacked compute power and data over the past two years, why hasn't it dominated the Chinese models? Some also questioned the actual contribution of JPMorgan to Kimi's revenue.

Some even sarcastically remarked that if this is the reasoning level of a "Muse Spark core member," they are starting to worry about Muse's model performance.

Muse Spark 1.2 is about to release its weights, introducing significant U.S. competitors like Kimi, DeepSeek, and Qwen. However, the leap from "having one more formidable opponent" to "Chinese models will be overwhelmed by compute power, and U.S. revenue will also be lost" is quite a stretch.