Hey everyone 👋 We’ve been exploring the AI video space for a while, and one thing kept bothering us — most models still struggle with consistency. You can generate impressive single frames, but once it turns into video, things often break: - flickering - identity drift - unstable motion So we started experimenting with ways to improve temporal consistency and motion realism. Recently, we saw something unexpected — the model we’re working with (HappyHorse-1.0) ranked #1 on the Artificial Analysis leaderboard for image-to-video (no audio), slightly ahead of models like Seedance, PixVerse, and Kling. That said, we don’t think leaderboard results tell the full story. What really matters is: 👉 Can creators actually use this in real workflows? 👉 Does it save time, or still feel like a demo tool? We’re still early, and not everything is fully polished yet — but we’d love to hear how you’re currently using AI video tools. - What’s still broken for you? - What would make you actually rely on one? Really appreciate any feedback 🙏

HappyHorse AI

AI video generator for consistent, high-quality results

AI video generator for consistent, high-quality results