Tag: Kimi

MiniMax: Charting Its Own AI Path

Explore MiniMax's strategic pivots in the Chinese AI landscape, focusing on product-model integration, overseas expansion, and technological advancements amidst intense competition.

MiniMax: Charting Its Own AI Path

MiniMax's All-In Bet: A Singular Strategy

MiniMax carves its own path in AI, prioritizing overseas growth and model-product integration. Can this focus on technology and monetization help it stand out amid intense competition from other AI Little Tigers?

MiniMax's All-In Bet: A Singular Strategy

Kimi Moonlight 30B 160B MoE Model

Moonshot AI unveils Moonlight a hybrid expert model with 30B and 160B parameters trained on the Muon architecture using 57 trillion tokens. It achieves superior performance and Pareto efficiency with a novel optimizer that doubles computational efficiency compared to AdamW making large language model training more accessible and sustainable.

Kimi Moonlight 30B 160B MoE Model