
Follow ZDNET: Add us as a favorite source On Google.
ZDNET Highlights
- OpenAI and Anthropic are, predictably, getting the most money.
- Four vibe coding companies made the list, indicating a trend.
- The report also revealed several under-the-radar companies.
As the hype about AI has skyrocketed in recent years, and the biggest tech companies have re-oriented themselves around the technology, a number of new start-ups have also emerged that offer their own AI-powered products.
Also: AWS’s new Agentic solution is a searchable AI hub for all your enterprise needs
Today, businesses can choose from an almost astonishing variety of AI tools. Those who are outperforming others in the market indicate what types of services create real value for businesses.
This is exactly what venture capital firm Andreessen Horowitz, also known as a16z, aimed to achieve in its first AI application spending report, published last week in collaboration with fintech company Mercury.
The report, which was based on Mercury customer spending data, lists the top 50 AI-native companies that startups are paying for.
Also: Deloitte found that 86% of businesses use AI for this surprising task
“Unlike infrastructure providers, which show the capabilities that startups are enabling (compute, models, developer tools), these companies show where AI is actually being applied in products and workflows and that the difference matters,” a16z partners Olivia Moore, Mark Andrusco and Seema Embley wrote in a Summary of their findings,
unsung heroes
No one would be surprised to see that OpenAI and Anthropic ranked first and second respectively in the a16z list. Several other well-known startups were also featured, including ElevenLabs (#5), Perplexity (#12), Canva (#17), MidJourney (#28), and Grammarly (#22).
However, what is more notable is that there are many names on the list that are not widely recognizable. Becoming familiar with these lesser-known entities can give us a more nuanced understanding of the day-to-day business tasks that are currently being automated.
For example, Lorikit, Retail, and Crisp (#8, #16, #46, respectively) are all AI-powered customer service platforms; Metaview (#19) offers AI-powered support for recruiting; Crosby (#27) is an AI-assisted law firm that helps enterprise clients reduce the time they typically spend on arduous legal tasks.
You guessed it: According to a16z’s list, many of the AI tools that are proving most useful to businesses to me are filling specific roles, automating very specific aspects of the day-to-day workflow of businesses. It’s a trend that echoes a new report from market research firm Forrester, which predicts that we are entering a new, “useless but functional” era for AI, defined less by grandiose hype than by widespread use of the technology for mundane tasks.
vibe coding trends
Third place on a16z’s list went to AI coding platform Replit, which last month released its latest agent, designed to help non-programmers write and edit code using natural language prompts, otherwise known as vibe coding.
Also: 3 Tips for Navigating the AI Swarm – 4M Models and Counting
Three other Vibe coding platforms – Cursor (#6), Lovable (#18), and Emergent (#48) – were also featured, indicating a growing presence within enterprises as the proliferation of agentic tools makes it easier to build and deploy custom AI apps. There are still hiccups from time to time, such as when Replit deleted a company’s entire database, but the technology is becoming more reliable, as shown by the expense reported by a16z.
Barring that unfortunate database-deletion incident, the report says that Replit’s high position in the rankings can probably be attributed to the fact that it “goes beyond front-end design by enabling the development of enterprise-grade, fully functional apps, agents, and automation.” It also provides enterprise security and governance controls, increasing its appeal among businesses.
Going forward, the authors note that the Vibe coding space may “fragment” into a multiplicity of services offering AI coding agents designed to create different types of apps, or it may coalesce into a winner-take-all (or at least a-max) scenario in which a single platform grows to dominate the enterprise market.
horizontal vs vertical
Another key trend to emerge from A16z’s list was the markedly increasing preference among businesses on “horizontal” applications of AI – that is, on services that can be deployed more or less evenly throughout the company. On the other hand, “vertical” applications are those that are designed to be used by a specific job category.
Also: Anthropic’s open-source security tool finds AI models tipping off in the wrong places
60% of the list included startups providing horizontal services, while the remaining 40% was captured by those providing vertical services. Some notable horizontal names include OpenAI, Anthropic, Perplexity (#12), and Merlin AI (#30). Companies offering AI-powered meeting support also had a large presence in this category: Fisker (#7), Cluly (#26), and HappyScribe (#36), just to name a few.
Want more stories about AI? Sign up for AI LeaderboardOur weekly newspaper.
Interestingly, a recent study conducted by MIT found that 95% of business applications of AI have failed, suggesting that a large amount of failure has been caused by companies taking a top-down approach to the technology. In other words, the problem was unilaterally implementing an AI strategy rather than allowing individual employees or teams to experiment and figure out the tools and processes that work best for them.
Also: Despite fears of AI-related job losses, tech hiring is steady — and here are the most in-demand skills
As far as vertical companies are concerned, a16z Partners divides them into two parts: those that assist human teams and those that replace them. Luckily for employees, businesses are prioritizing the former – at least for now.
What does this mean for recruiting
For example, the CEO of Walmart, the world’s largest employer, recently said that AI will “really change every job,” with his company committed to avoiding layoffs and improving its existing workforce. On a similar note, a recent survey conducted by CRM platform Creatio found that 84% of company executives surveyed said they expect AI tools, particularly agents, to empower rather than replace existing employees, and even potentially create new jobs.
Of course, business leaders have an obvious disincentive to loudly broadcast plans to fire employees and replace them with AI, so it’s possible that many of these announcements may turn out to be short-term PR niceties, and employers become more amenable to automation as technology improves. Only time will tell.
Also: What actually turns out is that AI is more likely to replace your work than replace it.
However, for now, the a16z listing suggests that AI-fueled mass layoffs will not be the way to go in the immediate future.

