alt Hacker NewsThe competition in text-to-video tools is heating up, but a key challenge remains: achieving desired results without exhausting resources. Runway, for instance, often consumes all your credits before producing something usable, even if you stick to their guidelines. Hailuo AI shows better consistency, while Sora Turbo sounds promising with potentially more mature generations. Progress is clear, but there’s still a way to go in perfecting these tools.
I've been thinking about your point regarding credit consumption with these tools. I started exploring this at https://liveimage.ai (an AI video avatar generator) with a low-res preview system first. I'm curious if you think this kind of approach would be more useful - generating quick low-res previews that use minimal credits, then only processing high-res versions once you're happy with the result? Seems like it could help avoid wasting resources on full quality generations that don't match what you're looking for. I often wonder if jumping straight to high-res without previewing is partly why people burn through credits so quickly. Would love to hear if you've encountered other tools taking this kind of approach.