A new A16Z report has seen that AI companies are actually paying startups

On Thursday, Andresen Horovitz released it First AI Spend Report In partnership with Fintech firm Mercury. Using transaction data from mercury, the report analyzes the top 50 AI-country application layer companies, which are spending the same money as the startups first published. Top 100 General AI Consumer Apps,

A16Z partners Olivia Moore and Seema Ambal say that data show companies are still adopting a series of various AI products for some tasks – and new apps are growing and falling very quickly.

“There is the spread of equipment,” Ambal said. “It is not around one or two in each category.”

The report also shows a lot of expenses on “human organizers” or “copillots” that can help promote productivity between the workforce, suggests that startups are not ready to shift to agent workflow completely.

“As the computer is used more than a mode and has more ability to create end-to-end agent flow, I think there will be shift, where we will see less copylots and more end-to-end agent tools,” said Ambal. “Especially as people are really eager to make them an attempt.”

Unexpectedly, Major Labs was dominated at the top of the list, OpenIAI took the top slot and anthropic at number 2. Vibe-coding tools also represented well, with answer at number 3 and cute at number 18. Cursor landed at number 6 and emerging at number 48.

When A16Z produced a uniform list for consumer habits, cute or much ranked much more than the answer on traffic because many people were using it to make projects. But startups are not spending as much money on cute as they are in part due to lack of enterprise features. But in the list, various companies suggest that there is a place for many different companies at once.

“This is an open question that is moving on vibe coding,” Moore said. “Does the space begin to consolidate, and a place becomes the best platform for a vibe code? Or is it a case where there will be four or five and Vibe coding companies that are really big businesses for various types of applications? We have not yet the answer.”

Moore was also surprised to see startups adopting consumer-oriented equipment such as Capcut and Midjourney.

“We are seeing that many of these (consumer) companies are moving rapidly and rapidly into the enterprise because they make delightful consumer equipment that people then adopt and use as individuals and bring them to their teams and workplaces,” Moore said.

Horizontal applications made at least 60% of the names in the list overall, and 40% included vertical applications. The most popular vertical software companies fell into three buckets: sales, recruitment and customer service. But the report also found that AI is progressing in many areas that the previous generations of startups fought to crack.

Moore said, “Probably what will happen before, eg, service firm or consultation is now software companies at AI age,” Moore said.

Ambal gave the crossbie as an example legally, which could quickly review a legal contract for a user, at a point, an in-house was rumored through a meeting, ideas and research with an in-house general lawyer. He just said, most devices are used to assist employees (such as co-pilots), which is in rapid decision making, rather than replacing the entire workforce and talent suit with automated workers (end-to-end agents).

“As the technique gets better, and we are actually able to build full agent colleagues, you will see that the mix is ​​far from the shift end-to-end agents and away from the co-pilots,” he said, “Later AI equipment can do a lot of work, such as outreach, faster than a human can.”

There were also a lot of notes in the list, such as Otter.aiRead AI, and HappyScribe – No option dominates. This is the same when he said that when he said that there is not a product that is still dominating the market; Rather, the startups are still “taste”, to see what equipment they like. It is also good for employees, which, with so many options, instead of using a size-fit-all product, which apps can help them do the best work, “can push up from top to bottom,” Ambal said.

The final major discovery in the report was a growing interview of consumer and enterprise businesses. People are bringing individual applications used at home to work, and people who have started companies are using their favorite personal applications to help creating their businesses. Earlier, there was a delimitation between the two: a set stack for what to use during the construction of the startup.

Ambal and Moore quoted Canva as an example: it is a popular consumer app with a large -scale venture audience. It took years to add an enterprise plan to Canva. But as it becomes difficult to distinguish cases of individual and enterprise use, companies are more inclined to mix.

“Your tam (total addressable market) is no longer one or the other, but you can sell in both,” he continued. He said that companies that manufacture these products may rapidly “professional”, which means building enterprise teams, such as Go-to-market, sales and support, so they can start selling and acquiring enterprise revenue as soon as possible on the basis of individual consumers.

Moore and Ambal are expecting to change the list quickly in the coming years. Older companies are now launching AI features to remain relevant and accessible, while new entrances come with new ideas.

“Inheritance players, almost mean of heritage, ‘what was 12 months ago,” said Ambal. “If we pull it again in 12 months, will the same note -taking app be there too? Or a new set?”

This piece was updated to reflect the note -taking app.