Building a viable pricing model for generative AI capabilities can be challenging

October, box unveiled a new pricing approach for the company’s generative artificial intelligence capabilities. Instead of a flat rate, the company designed a unique consumption-based model.

Each user can obtain 20 points per month, which is applicable to any number of AI tasks. Up to 20 new events can be added, and one point is charged for each task. After that, people can use the 2,000 additional points provided by the company. If a customer exceeds this limit, they should discuss purchasing additional points with a salesperson.

Box CEO Aaron Levie explained that this approach provides a way to charge based on usage because we know some users will take advantage of AI capabilities more than others, while also taking into account the use of OpenAI, which the company is using. The cost of the API is its underlying large language model.

Meanwhile, Microsoft has opted for a more traditional pricing model, announcing in November that it would charge $30 per user per month to use its Copilot capabilities, on top of the cost of a normal monthly Office 365 subscription. Fees vary by customer.

While it has been clear over the last year that enterprise software companies will build generative AI capabilities, during a panel discussion on the impact of generative AI on SaaS companies at Web Summit in November, the co-founder and chief executive of communications API startup Nylas Christine Spang, Technology Officer: Manny Medina, CEO of sales enablement platform Outreach, spoke about the challenges SaaS companies face when implementing these capabilities.

First, despite the hype, generative AI is clearly a huge leap forward, and software companies need to find ways to integrate it into their products, Spang said. “I wouldn’t say 10 out of 10 the hype is in line with reality. [current] Realistic, but I do think there’s real value there and what’s really going to make an impact is how people take this technology and connect it to other systems, other applications, and leverage it in different use cases to drive real value,” she says.