// news · multimodal2026-06-12source: artificial analysis / pixflow / lovart

Kling v3 holds text-to-video leaderboard leadership at arena score 2031 — China-built video generation outranks LTX-2 Fast and Happy Horse 1.0 going into late June

Kling v3 leads the Artificial Analysis text-to-video leaderboard with an arena score of 2031, followed by LTX-2 Fast (1930) and Alibaba's Happy Horse 1.0 (1893). The top-3 ranking is now China-built across all three positions, with the closest US-frontier entries (Runway Gen-4.5, Pika 2.5) sitting outside the top tier on quality benchmarks while remaining the procurement choice for marketing buyers.

The substantive piece is the segmentation. The arena-score ranking measures quality on standardized prompts; the procurement market segments by use case. Runway leads marketing-buyer share through Gen-4.5's reference-image control and character consistency; Kling leads quality-leaderboard share. For the enterprise buyer choosing a video stack, the answer is increasingly to license multiple — Kling for quality-tier work, Runway for brand-consistency production, Veo 3.1 for photorealism, Pika for stylized b-roll.

The competitive frame is that Alibaba's Happy Horse 1.0 leaderboard reset is now the third position rather than first — Kling v3 reclaimed the top spot. The three-month leaderboard volatility tells the structural story: video generation is the fastest-iterating multimodal category, with China-built models holding the top three slots as of June 2026.

See our analysis →

LLM Stats — Best AI for Video Generation in 2026 — Ranked by Blind Human Votes → · Pixflow — Best AI Video Generators in 2026 - Free & Paid Ranked → · Lovart — Best AI Video Generators in 2026: Sora 2 vs Runway vs Pika Labs →