Kling v3 holds text-to-video arena leadership at 2031 score — Chinese physics-understanding edge widens against Veo and LTX-2 Fast
Kling v3 holds the text-to-video arena leaderboard at score 2031, followed by LTX-2 Fast (1930) and Happy Horse 1.0 (1885). The China-platform physics-understanding edge (hair, liquids, fabric motion) plus multi-shot storyboarding with native audio sync is widening Kling's lead at the standalone-platform tier even as Google Veo 3.1 dominates product-integration deployments.
The substantive piece is the arena-score persistence. Blind human-vote leaderboards are the most credible AI-video evaluation methodology because they sidestep benchmark-gaming pressure. Kling v3's 2031 score holding through mid-June across the second week of evaluation means the platform's quality edge is genuine rather than launch-momentum. The 100-point gap to LTX-2 Fast at 1930 is structurally meaningful.
The competitive frame for the three-tier video-generation market is hardening. Google Veo 3.1's product-integration deployment serves the embedded-AI surface; Kling v3 serves the standalone-platform-with-global-enterprise-tier; OpenAI's Sora wind-down (web April 26, API September 24) leaves the standalone-platform-western tier to Runway, Pika, and Luma. Each tier now has a clear procurement-decision path.
LLM Stats — Best AI for Video Generation in 2026 — Ranked by Blind Human Votes → · AI/ML API — Best AI Video Generators 2026: Veo 3.1, Kling, Sora 2, Seedance & More Compared → · Atlas Cloud — Best AI Video Generation Models in 2026: Complete Comparison →