AI-generated video is becoming easier to produce, but cost, speed and brand consistency remain significant barriers for advertisers creating large volumes of creative. Creatify Labs is targeting those constraints with Boreal, a text-to-video and image-to-video model that generates short clips in near real time and is priced at one cent per second of finished video.
Generating an AI video can take minutes, cost several dollars and still produce a result that changes the product, person or interface shown on screen. For advertisers producing dozens or hundreds of creative variations, those limitations can make experimentation expensive and operationally slow.
Creatify Labs is taking a different approach with Boreal, an AI video generation API designed specifically around advertising production. The company says the model can generate a five-second clip in roughly five seconds, while pricing is set at $0.01 per second of finished video.
That means a five-second generation costs about $0.05, according to Creatify. The pricing model also removes some of the uncertainty associated with credit- or token-based AI services because advertisers can calculate the generation cost directly from the duration of the finished video.
Speed is the other major part of the proposition. Creatify says Boreal operates at approximately 1:1 real-time generation, meaning one second of video can be produced for approximately one second of GPU processing time.
For advertising teams, that changes the economics of creative iteration. A campaign team testing multiple hooks, products, scenes or creator variations can generate and evaluate more alternatives without waiting for lengthy rendering queues.
Creatify says Boreal is built on an open-source video generation model capable of generating audio and video jointly. The company then applied a post-training process using advertising footage, with the goal of improving the characteristics that matter in commercial creative.
One of those characteristics is product fidelity.
General-purpose video models can introduce subtle but damaging changes to packaging, logos, product shapes, creator identities or application interfaces. Those errors may be tolerable in entertainment or experimentation, but they can undermine an advertisement where the product itself needs to remain recognizable.
Creatify’s testing focuses heavily on that issue. In blind comparisons involving 40 production cases, evaluators preferred Boreal over the underlying untrained model in 81% of decisive comparisons, according to the company’s report. The test included 25 preferences for Boreal, six for the base model and nine ties.
The company also reports that Boreal performs competitively against a leading closed video model on the perceptual measures it considers important for advertising, while generating video substantially faster and at a lower benchmarked cost.
Those comparisons should be treated as vendor-reported rather than independent industry benchmarks. Nevertheless, the underlying strategy is notable: instead of competing solely on visual quality, Creatify is optimizing around advertising-specific reliability, generation economics and throughput.
The economics become more apparent at scale. Creatify estimates that producing a storyboard containing twenty five-second clips costs around $0.25 on Boreal and can be completed in under two minutes. Its benchmark comparison puts the same workflow at roughly $12 and more than an hour on a leading closed model.
The model is not positioned as universally superior. Creatify says Boreal is currently strongest on short clips and relatively simple scenes, while performance is closer to the base model for single-person talking sequences.
That caveat matters as advertisers evaluate AI video platforms. Creative teams need more than fast generation; they need predictable outputs, brand controls and consistent characters and products across variations.
Boreal’s API-first approach puts those requirements directly into the advertising production workflow. If the model can maintain its reported speed and fidelity as workloads grow more complex, inexpensive real-time generation could make continuous creative iteration a more practical part of performance advertising.
Market Landscape
AI video generation is moving from experimental content creation toward commercial production. Models from major AI companies and emerging specialist providers are competing on visual quality, motion, consistency, controllability and generation speed.
For advertisers, however, the economics are different from those of consumer video creation. Large campaigns can require thousands of variations across products, audiences, languages and placements. At that scale, even small differences in generation cost and turnaround time can materially affect creative testing workflows.
Boreal’s one-cent-per-second pricing and real-time generation target that specific problem. Its advertising-focused post-training also reflects an emerging specialization within generative AI: models optimized not simply for creating attractive video, but for maintaining commercially important details.
The competitive landscape includes closed models from major AI providers as well as open-source video-generation systems that developers can customize. Creatify’s strategy combines an open foundation with advertising-specific post-training and an API designed for production use.
For enterprise advertising teams, the decision will ultimately depend on more than benchmark quality. Brand safety, intellectual-property considerations, content controls, API reliability, scalability and consistency across large creative libraries will become increasingly important as AI-generated video enters mainstream campaign production.
Top Insights
- Creatify Labs’ Boreal generates text-to-video and image-to-video content at one cent per second, targeting high-volume advertising creative production.
- The company says Boreal generates five-second clips in roughly five seconds, enabling near-real-time iteration for advertisers and creative teams.
- Boreal uses advertising-focused post-training to improve product fidelity, creator consistency, interface legibility and other requirements of commercial video.
- Creatify’s internal testing found reviewers preferred Boreal over its underlying open model in 81% of decisive comparisons across 40 production cases.
- The launch highlights a shift in AI video competition toward generation economics, throughput and advertising-specific reliability alongside visual quality.
Get in touch with our Adtech experts
