Camera Intelligence Brings Google's Omni and Runway's Aleph to Caira, the First Mirrorless Camera System with Generative Video Editing via World Models

London and New York — September 2nd, 2026 — Camera Intelligence, the AI camera startup building Caira, an AI-native Micro Four Thirds mirrorless camera for creators and brand owners, today announced its Generative Video Editing feature — “Nano Banana for video editing”. Powered by Google’s Gemini Omni 1.1 Flash (the “any-to-any” world model first introduced at Google I/O 2026) and Runway’s Aleph 2.0, Camera Intelligence is the first company to embed world models directly into a mirrorless camera system.

The new capability lets users apply visual and camera-motion effects in the moment to footage they have already captured with Caira, using natural language — all through the Caira app which controls the camera.

The feature has been built in response to customer demand following a landmark first year for Caira, whose December 2025 Kickstarter campaign passed its funding goal in under 24 hours and raised over $460,000. During the campaign, Camera Intelligence first announced generative photo editing via Google’s Nano Banana model. Having delivered the majority of Kickstarter units — with remaining units continuing to be delivered every month — the company surveyed 111 backers about what should come next: 77% of respondents said generative video editing would be useful to them, directly shaping the product roadmap.

Editing via world models

Unlike earlier generative models that edit one frame at a time, a world model builds an internal understanding of the scene itself — its geometry, lighting, materials, and motion. Gemini Omni is physics-aware and remembers the scene across an entire clip, so footage can be relit or extended realistically, while people, objects, and camera movement stay consistent from frame to frame.

Caira’s Generative Video Editing feature launches with Relight, Reframe, Camera Motion, and Set Extension. Editing works through a curated library of templates, pre-designed by the Camera Intelligence team in collaboration with feedback from their community. Users can refine and control a template further with conversational editing — plain-English instructions interpreted within Caira’s ethics guardrails. Generations are rendered in 720p or 1080p and take between 1–3 minutes depending on resolution.

There is no text-to-video function: every edit begins from footage the user has captured; a scene cannot be generated from nothing. Generated results are saved as separate, clearly named files, and are watermarked with Google’s SynthID, a tool to signal provenance and identify AI-generated content. Original mp4 files are never modified, Caira’s on-camera image pipeline remains non-generative, and the new feature is entirely optional.

By embedding this technology directly at the point of capture rather than confining it to post-production software, Camera Intelligence enables rapid previsualization and prototyping during pre-production, allowing creators to expand their creative possibilities and evaluate options before committing to specialized VFX work. With Generative Video Editing, a Caira user can shoot a spec sequence in the morning and present three treatments the same day — then, when the pitch lands, shoot the real production with a full crew. Camera Intelligence positions the feature as a way to rapidly prototype, not to replace production.

Built with the community

“Last year we were the first camera to edit photographs with generative AI, and we learned a great deal from the conversation that followed,” said Vishal Kumar, CEO and co-founder of Camera Intelligence. “This year, after surveying our backers, 77% said that generative editing capabilities for video would be useful to them, giving us a clear impetus to bring these ideas to life.”

The feature is designed to close the gap between capture devices and editing tools. Traditional filmmakers and VFX artists may prefer their existing workflows; Caira’s solution is aimed squarely at solo creators and marketers who want creative options available to them in a more straightforward and easy way — more creative options in the moment, to help visualise and illustrate what is in their head before shooting.

“We understand there are strong emotions around AI in photography and filmmaking right now, and we want to be cognizant, sensitive and respectful towards artists and their craft,” Kumar added. “A lot of our audience are overwhelmed by more pro tools, but extremely excited and intrigued by these technologies. Listening to our community ensures we build products that are innovative, fresh, and useful for them. We will monitor usage closely and continue building features that genuinely support our community’s work.”

Availability and pricing

Generative Video Editing — alongside Generative Image Editing and a yet-to-be-announced feature — will be part of Caira Studio, a premium software offering launching in Autumn 2026. Details on pricing will be available soon. Generative Video Editing will be available in the US and UK.

A private beta opens today for existing Caira customers, along with 50 places for working filmmakers. Express interest at cameraintelligence.com/generative-video-editing or via team@cameraintelligence.com.

Ethical AI for creative empowerment

Camera Intelligence integrates advanced generative AI to enhance creative expression — and does so within the framework the company adopted after listening to its community, creators, and customers over the past year. All on-camera image processing remains strictly predictable, repeatable, and free from generative AI. Generative Video Editing, like all of Caira’s generative features, is fully optional and exists exclusively after footage has been captured.

The Camera Intelligence Ethical AI Charter:

• Customer footage and images are never used for model training.
• There is no open prompt box: generative editing operates within curated, defined categories.
• Every generated output is a separate, clearly named file with provenance metadata; originals and RAW are never modified.
• The on-camera image pipeline is non-generative — predictable and repeatable.
• All generative features are optional and off by default.
• Guardrails prohibit altering a person’s skin tone, ethnicity, or fundamental facial features.
• Camera Intelligence uses pre-vetted, commercially available models, and expects to adopt licensed, opt-in models as they mature.
• All deployed technology adheres to Google’s Generative AI Prohibited Use Policy.

Note to editors

No paid creator or media partnerships are associated with this launch; review units are provided on a loan basis where creators and editors have full creative freedom. The press kit includes the survey methodology, hi-res images, B-roll, and an FAQ. Interviews with Vishal Kumar are available at the 2 September launch event at Digital Catapult, London, or by arrangement: team@cameraintelligence.com.

About Camera Intelligence: Camera Intelligence builds the AI-native camera. By integrating advanced AI directly into a mirrorless camera system that connects to smartphones, Camera Intelligence puts intelligence in creators’ hands at the point of capture — the hardware is how the intelligence reaches creators, filmmakers, and e-commerce businesses.

Contact:
Vishal Kumar, CEO & co-founder
team@cameraintelligence.com
www.cameraintelligence.com