Multiple outlets report that ByteDance is working on a very large “mega” AI model that is nearing the capabilities of Anthropic’s flagship model, Mythos, according to a Financial Times report. The reporting indicates that ByteDance’s effort is focused on training and scaling a new system designed to compete at the top end of the large language model field. While the sources do not provide full technical specifications in the excerpts provided, they describe the effort in terms of model size and performance benchmarks relative to Mythos. The articles also frame the work as part of broader industry competition to develop leading frontier AI systems. No details are included here about timelines for release, deployment, or regulatory approvals, nor about whether ByteDance plans to make the model available publicly or restrict it to internal or partnered use. Overall, the information points to accelerated model development at ByteDance intended to match or closely approach Anthropic’s most prominent large model offerings.