Teaching machines to make video that feels alive. Frontier AI models from the research lab at Creatify AI. Live on fal.
- Tweets 146
- Following 5
- Followers 344
- Likes 229
Pinned Tweet
Introducing Boreal-H3 — a video model built for ads and our next step toward recursive self-improvement in video generation.
A good-looking video isn’t enough. The product has to stay the same. The actor has to stay the same. The label has to be right. And the action in the brief actually has to happen.
So we post-trained MiniMax H3 specifically for advertising.
But this isn’t a one-off SFT or LoRA fine-tune. We built a closed-loop system that learns what to improve next.
Human-calibrated evaluation diagnoses failures and guides the next intervention: targeted data collection, reinforcement learning, or inference optimization. When the feedback is unreliable, we revise the evaluator or reward—not just the generator.
Every experiment feeds into shared memory, informing the next training decision. The model improves, and so does the process that produces its successor.
The results:
→ 85.3% reference fidelity — highest among the frontier video generation models we evaluated
→ Brief success: 28% → 50%
→ Identity match: 83% → 94%
→ Visible defects per clip: down 70%
→ Generation time and estimated cost: down 20%
Boreal-H3 doesn’t just make better-looking video. It makes more usable ads.
Credit to the @MiniMax_AI team for the foundation we’re building on.
This launch is a checkpoint, not the finish line. We’re building more than a better video model. We’re building a system that learns how to make the next one better.
91
230
203
757,927
wooow
I've made this video in Creatify for Flowmapp in ONE shot!
Opus 5.5 + Boreal-H3
i gave the agent the website and it did the rest
(a year ago making something like this would've been a whole video production process)
wtf happens to creative industry if this gets 10x better from here?
Introducing Boreal-H3 — a video model built for ads and our next step toward recursive self-improvement in video generation.
A good-looking video isn’t enough. The product has to stay the same. The actor has to stay the same. The label has to be right. And the action in the brief actually has to happen.
So we post-trained MiniMax H3 specifically for advertising.
But this isn’t a one-off SFT or LoRA fine-tune. We built a closed-loop system that learns what to improve next.
Human-calibrated evaluation diagnoses failures and guides the next intervention: targeted data collection, reinforcement learning, or inference optimization. When the feedback is unreliable, we revise the evaluator or reward—not just the generator.
Every experiment feeds into shared memory, informing the next training decision. The model improves, and so does the process that produces its successor.
The results:
→ 85.3% reference fidelity — highest among the frontier video generation models we evaluated
→ Brief success: 28% → 50%
→ Identity match: 83% → 94%
→ Visible defects per clip: down 70%
→ Generation time and estimated cost: down 20%
Boreal-H3 doesn’t just make better-looking video. It makes more usable ads.
Credit to the @MiniMax_AI team for the foundation we’re building on.
This launch is a checkpoint, not the finish line. We’re building more than a better video model. We’re building a system that learns how to make the next one better.
Paid partnership (ad)
4
4
21
1,546
Creatify Labs retweeted
Opus 5.5 + Boreal-H3 is getting seriously good.
This entire ad came from one ask.
One shot. No stitching together 15 generations. No babysitting every scene until something finally looks usable.
Creatify’s Ad Agent actually turned the brief into the ad.
That’s a much bigger deal than another flashy AI video demo.
Introducing Boreal-H3 — a video model built for ads and our next step toward recursive self-improvement in video generation.
A good-looking video isn’t enough. The product has to stay the same. The actor has to stay the same. The label has to be right. And the action in the brief actually has to happen.
So we post-trained MiniMax H3 specifically for advertising.
But this isn’t a one-off SFT or LoRA fine-tune. We built a closed-loop system that learns what to improve next.
Human-calibrated evaluation diagnoses failures and guides the next intervention: targeted data collection, reinforcement learning, or inference optimization. When the feedback is unreliable, we revise the evaluator or reward—not just the generator.
Every experiment feeds into shared memory, informing the next training decision. The model improves, and so does the process that produces its successor.
The results:
→ 85.3% reference fidelity — highest among the frontier video generation models we evaluated
→ Brief success: 28% → 50%
→ Identity match: 83% → 94%
→ Visible defects per clip: down 70%
→ Generation time and estimated cost: down 20%
Boreal-H3 doesn’t just make better-looking video. It makes more usable ads.
Credit to the @MiniMax_AI team for the foundation we’re building on.
This launch is a checkpoint, not the finish line. We’re building more than a better video model. We’re building a system that learns how to make the next one better.
Paid partnership (ad)
2
8
38
9,082
Introducing Boreal-H3 — a video model built for ads and our next step toward recursive self-improvement in video generation.
A good-looking video isn’t enough. The product has to stay the same. The actor has to stay the same. The label has to be right. And the action in the brief actually has to happen.
So we post-trained MiniMax H3 specifically for advertising.
But this isn’t a one-off SFT or LoRA fine-tune. We built a closed-loop system that learns what to improve next.
Human-calibrated evaluation diagnoses failures and guides the next intervention: targeted data collection, reinforcement learning, or inference optimization. When the feedback is unreliable, we revise the evaluator or reward—not just the generator.
Every experiment feeds into shared memory, informing the next training decision. The model improves, and so does the process that produces its successor.
The results:
→ 85.3% reference fidelity — highest among the frontier video generation models we evaluated
→ Brief success: 28% → 50%
→ Identity match: 83% → 94%
→ Visible defects per clip: down 70%
→ Generation time and estimated cost: down 20%
Boreal-H3 doesn’t just make better-looking video. It makes more usable ads.
Credit to the @MiniMax_AI team for the foundation we’re building on.
This launch is a checkpoint, not the finish line. We’re building more than a better video model. We’re building a system that learns how to make the next one better.
91
230
203
757,927
6/
A concrete example:
If the model starts drifting on product packaging, the answer isn’t simply “train more.”
The system can identify the failure, collect more packaging and product-handling data, curate it, train a new candidate, and test whether fidelity improves without regressing elsewhere.
And if the evaluator itself is noisy, we fix the evaluator first.
Every experiment becomes memory for the next intervention.
That’s what we mean by toward recursive self-improvement in video generation.
1
12
598
Try Boreal-H3: app.creatify.ai/tool/ai-tool…
Try Ad Agent: app.creatify.ai/ad-agent
14
589
Introducing Aurora 2.0 —our best avatar model 🔥
One photo. Your exact voice. Up to 60s of expressive video. Sharper lip sync. Natural gestures. 2× faster than Aurora. $0.07/s.
🏆 Highest overall score in our avatar benchmark.
Build UGC ads, podcasts, product explainers, and character videos. Swap the voice track to test a new hook, change the offer, or take your campaign into another language.
9
5
28
1,886
More about Aurora 2.0:
labs.creatify.ai/models/auro…
1
1
3
224
Wait, guys, I think this play is about us🙊👀
We built our own video model.
Then we rebuilt our most-used tool on it.
URL-to-Video now runs on Boreal.
Same flow. A whole new engine.
With a single URL, you can:
> Create ads in viral formats we know win
> Customize scripts and tones
> Choose from a huge library of animated and realistic avatars
9
16
580
Big fashion houses have always had the budget for films like this. Now smaller brands have the tool.
With Boreal, you get the same runway energy.
Comment "RUNWAY" for the prompt.
11
1
17
440
Try Boreal on fal: fal.ai/models/creatify/borea…
1
3
138
Try Boreal on fal: fal.ai/models/creatify/borea…
15
Your next pet ad doesn't need treats to bribe the talent.
Swap the boots for your own product.
When an ad costs 1¢ per second and it's too cute to ignore, you've already won.
10
14
782
Try Boreal on fal: fal.ai/models/creatify/borea…
1
102