Camera Intelligence is a startup which has decided to change cameras – it might not be for everybody, but the company thinks its Caira camera might sell to ten times as many to creatives than the entire mirrorless market. Until today, I wasn't sure why.
Last year we covered the Kickstarter for the Caira, in which Camera Intelligence CEO Vishal Kumar unveiled a Micro Four Thirds camera system which – he says – attaches to a phone and bridges the gap between the photographic aspirations of people who buy the 'iPhone Pro' series (of whom there are 100 million every year) and the desire for simplicity and easy of use which limits real camera purchase to 10 million a year.
The result is a piece of hardware with a SnapDragon processor and a touchscreen that has similar power to a smartphone, and a MFT 4:3 image sensor.
A subscription will be needed for the use of the Nano Banana feature, but I was able to test it out in person and I can see the appeal.
Some of the most impressive effects demonstrated were video – from a simple instruction and a short clip the camera could create a whole clip with consistent style – for example by telling it: "orbit this statue like a drone."
The device, called an "AI-native Micro Four Thirds" was backed with $459,000 (670 backers) and the devices are now shipping as they are delivered.
As of today's event, new generative features on the device included Nano Banana for video editing, which is part of the Gemini Omni 1.1 Flash "any-to-any" model introduced at Google I/O as well as Runway's Aleph 2.0 which adds "real-world understanding" (helping it create effects and even apparent 3D motion to clips which have already been captured).
The ultimate goal for Camera Intelligence is to eliminate the three stage workflow that photographers presently work through, especially creators in a hurry. By eliminating the memory card and on-computer edit, and instead achieving video and photo editing in device (or, at least apparently), before sharing directly through your phone's usual means.
Now of course we have seen some of these features – tone adjustments and the like – in other cameras, but I was astounded by the demonstration I saw, inspired by the Harry Potter movies, no doubt – in which a photo of St. Pancras Station in London was turned into a drone-style flypast.
This was made possible with the new AI tools that create and extend video from that you shoot in device. The high-speed integration of this feature (it's only been days since it was announced) gives an idea of how this camera works with software agents first and foremost.
The problem that the company has set out to solve is not that there aren't professionals looking for professional workflows, but that there is a camera workflow that hasn't changed in 30 years, which puts creators spending 20% of their time creating on device, and 80% editing.
By attaching a MFT system to a phone, and Google's "Edge" TPU on device, as well as a Wi-Fi connection to the phone, everything is there for computational photography not just in the 'iPhone sense', but LLM instructions. This combines to form what Kumar called a "Vertically integrated platform."
So, can Camera Intelligence find a market for cameras that has eluded the rest of the world? I don't see why not myself. After all, many people are happy to buy phone accessories (especially if they'll last multiple generations), and many find traditional camera system menus somewhat bewildering.
Being able to interface with your serious camera via a phone-like touchscreen app seems to bridge a lot of that gap (and might do so even more for younger users with no prior experience of cameras).
There will be frustrations. The Nano Banana – and any other AI image generator which is integrated onto the device at a later date – will need to be paid for by subscription.
There are also, in the form I tried, limitations which seem a good idea – AI content is generated from material you take, and cannot do things to change people's appearance.
In the world where you need to publish 3-4 times a week to stay relevant, a device with AI on hand could help out. I can only imagine more need for features like this. Of course, if features are added by a simple app update, is there a line that can be drawn? I'm not sure, but I am more than intrigued.
The final hardware and software are expected in Fall 2026, with generative video only available in the USA and UK at first.
