"Brute Force Yields Miracles": Report: ByteDance Discusses Training Model with Over 5 Trillion Parameters to Secure Leading Position
ByteDance is discussing training a large model with over 5 trillion parameters, aiming to secure an industry-leading position through its "brute force yields miracles" strategy. The initiative is led by Xiang Liang, head of the Seed Foundation, and is currently in its early stages with no final release confirmed. This move is seen as a breakthrough from the existing catch-up mode but also carries high risks
It is understood that ByteDance is discussing training a model with a parameter scale exceeding 5 trillion. This would surpass Alibaba's Qwen 3.8-Max (2.4 trillion parameters) and Moonshot AI's K3 (2.8 trillion parameters), making it the largest known model in China by parameter count. Generally, larger model scales often correlate with higher levels of intelligence. However, this plan is still in its early stages and does not guarantee the model's eventual release.
The new model will be led by Xiang Liang, head of the Seed Foundation, in collaboration with Shen Ke, who is responsible for pre-training data for large language models. To support this, Seed is reorganizing its structure, clarifying responsibilities, and allocating resources.
Several ByteDance insiders told us that many teams within Seed have been continuously reflecting during the first half of the year. This is one of the reasons ByteDance is discussing training a model with over 5 trillion parameters: rather than continuing to catch up on existing scales, it prefers to push the parameter scale to several times that of its peers at once to strive for a leading position. "It's like a gamble," said a source close to Seed.
Since its establishment, ByteDance has advocated the principle of "brute force yields miracles": when entering a new field, it typically establishes more than one team, invests substantial resources, and advances multiple projects simultaneously. (LatePost)
Risk Warning and Disclaimer
The market involves risks; investment should be approached with caution. This article does not constitute personal investment advice, nor does it take into account the specific investment objectives, financial status, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investors bear full responsibility for their own decisions.
