Kira 5.7 Nova & Ultra Video Model Update
Released: Aug.25.2026
Kira 5.7 brings a major video model update to Nova and Ultra.
Both models now generate better video, and Ultra does it a lot faster than anything else out there.
What's New in Kira 5.7
Updated Video Models for Nova and Ultra
Kira 5.7 brings major upgrades to video generation across both Nova and Ultra. The new video models deliver stronger generation and editing capabilities, with better overall results and improved prompt following.
Stronger video generation and editing
Better prompt understanding and consistency
Improved overall video quality
Automatically available for all new generations, no settings to change
Faster Generation on Both Models
We’ve also significantly reduced video generation times across Nova and Ultra.
Ultra
Most videos finish in around 1 minute
Built for the best results with minimal waiting
Nova
Most videos finish in around 5 minutes
A faster option for everyday video generation
Adjusted Video Pricing
Alongside the model and speed upgrades, we’ve adjusted the credit cost for video generation:
Nova: 15 credits per second of video
Ultra: 25 credits per second of video
Nova remains the more cost-efficient option, while Ultra is designed for users who want our strongest video capabilities and fastest generation speeds.
Here are some example for you to see the difference:
How It Works
Open Kira and start a video generation
Pick your model: Nova (15 credits/sec) or Ultra (25 credits/sec)
Describe your video and generate: Ultra finishes in about a minute, Nova in about five
Why This Matters
Waiting is the worst part of video generation. With Ultra now finishing in about a minute, you can iterate on ideas in near real time, try a prompt, see the result, adjust, and go again.
Two models, faster speeds, one simple choice, pick what fits the moment.
What's Next
We're continuing to push video generation forward:
Further speed and quality improvements across both models
More video creation options and controls
As always, your feedback helps shape what comes next.
