Sources: ByteDance is pre-training an AI model with up to 10T parameters, roughly 3x Kimi K3 and potentially larger than estimates for Anthropic's Mythos 5 (Financial Times)
The reported run is already in pre-training, with a stated ceiling of 10 trillion parameters.
Techmeme cites the Financial Times saying ByteDance is training a model roughly three times larger than Moonshot’s Kimi K3. The note says that scale would also exceed an 8 trillion-parameter estimate for Anthropic’s Mythos 5. Techmeme lists Reuters and The Decoder among follow-on coverage, plus social posts reacting to the reported size. Techmeme's note
Techmeme cites the Financial Times saying ByteDance is training a model roughly three times larger than Moonshot’s Kimi K3. The note says that scale would also exceed an 8 trillion-parameter estimate for Anthropic’s Mythos 5. Techmeme lists Reuters and The Decoder among follow-on coverage, plus social posts reacting to the reported size. Techmeme's note
score 6