Is AI worth paying four times as much for a Canva subscription?
Design software service Canva is increasing membership prices for users globally, with some U.S. users reporting on X that the cost of five-person team subscriptions has gone up from $120 per year to $500 per year. Canva reportedly told users in an email that it would discount the higher price by 40% for the next year, bringing it down to $300 — still a $180 increase.
The company is worth $26 billion, has 185 million global monthly users, and may be preparing to go public.
Canva is justifying the price hikes by touting its new AI features.
“Our suite of products has grown significantly in the last couple of years with the launch of new offerings like the Visual Suite and Magic Studio,” a Canva spokesperson said in a statement.
Visual Suite includes features like AI-powered writing, slideshows from a prompt, and AI turning documents into websites; Magic Studio uses AI to turn written concepts into social posts, videos, and slideshows.
Even if the new features got an AI, subscribers took to social media to express frustration over the price increase. One user on X noted that small businesses could feel the impact of the pricing change the most.
Canva renewed a multi-year partnership last month to give its customers access to Getty images and videos and also acquired AI startup Leonardo.ai in July to bolster its AI offerings. Leonardo’s full 120-person team joined Canva’s 4,000-person workforce as part of the acquisition.
“With every single AI feature that we introduced into Canva, we really think it’s helping to cut down that time between A and B of getting the idea into an actual product,” Canva CEO Melanie Perkins said in July.
Canva CEO Melanie Perkins. Credit: David Paul Morris/Bloomberg via Getty Images
Incorporating AI into a business isn’t cheap and comes with a significant environmental cost, too.
AI models currently cost over $100 million to develop, with the price tag only expected to rise over the next couple of years. AI researchers also estimate that it takes 30 times as much energy to have AI generate text instead of taking it directly from the source.
