Moonshot AI

Latest AI news, models and releases from Moonshot AI. ['GLM 5.2', 'GLM-5.2', 'Hermes', 'K2', 'K2.6', 'K3', 'K3-mini', 'K3-MoE', 'Kimi', 'Kimi 2.6', 'Kimi-2.6', 'Kimi 3', 'Kimi 3T', 'Kimi Hosted Agent', 'Kimi k1.5', 'Kimi K1.5', 'kimi-k2', 'Kimi K2', 'kimi-k2-0711-preview', 'Kimi K2.5', 'Kimi-K2.5-1T-A32B', 'Kimi K2.6', 'Kimi K2.7', 'Kimi K2.7 Code', 'Kimi K2 Thinking', 'Kimi K3', 'Kimi-K3', 'Kimi K3.5', 'Kimi K4', 'Kimi K5', 'Kimi Linear', 'Kivine', 'Krea 2', 'Krea 2 Turbo', 'Mamba', 'Mamba-2', 'Mamba-3', 'MiniCPM-RobotManip', 'MiniCPM-RobotTrack', 'Minimax']

Official site ↗

Business & Market 🇨🇳

2026: Why Capital Prefers AI Apps That Can Make Money

VC benchmarks changed: DAU no longer the only metric; high gross margin and retention are key. Haiyi, a Chengdu-based AI app company, raised over 100 million yuan in Series B led by Visual China and others, with 40%+ gross margin, 60%+ renewal rate, and $60 ARPPU. Its three product lines (SeaArt, MoreShort, SeaSoul) share a modular capability system enabling rapid scaling and sustainable commercialization.

Moonshot AI

Haiyi

36Kr27.07 · 13:02

Models 🇨🇳

Comparison of Three Chinese Models: Kimi K3, GLM-5.2, DeepSeek V4 Pro — Who Has the Best Coding Skills?

On the OSCHINA platform, a comparison of three Chinese AI models: Kimi K3, GLM-5.2, and DeepSeek V4 Pro was conducted. Their coding abilities were evaluated. The results show which model leads in coding.

Moonshot AI

DeepSeek

DeepSeek (GNews)27.07 · 12:02

Show more ↓