Introduction

Where startups really spend their AI budgets - a peek into which tools win the biggest share and why “AI-native” apps are reshaping work.

What’s the problem it solves?

The report uncovers how startups actually use AI - not through hype, but real spending data. It shows which tools are taking over company budgets, how they change team structures, and which roles are turning into AI-driven skills.

Quick Summary

The report, created by investors at a16z with data from Mercury’s 200,000 startup customers, tracks where early-stage companies spend their money on AI apps. It’s not about infrastructure or GPUs - it’s about real software that teams buy and use every day.

Most spending (60%) goes to horizontal tools, like OpenAI, Anthropic, and Notion, which boost productivity across all roles. Meeting assistants like Fyxer, Otter AI, and Read AI are popular, showing how everyday work is being automated. Creative tools like Freepik, ElevenLabs, Canva, and Capcut also dominate - AI has made creativity everyone’s job, not just for designers.

Vertical tools (40%) focus on specific jobs. Some help humans do their work better, like customer service or recruiting apps, while others replace whole teams. Examples include AI engineers (Cognition) or legal agents (Crosby Legal). The trend points toward startups “hiring” AI employees instead of people.

The fastest-rising field is “vibe coding” - tools like Replit and Cursor that let users build their own AI-powered apps and agents. Replit stands out as the most used by companies, showing how fast software creation is becoming automated.

Finally, many of the most popular AI tools began as consumer apps and are now becoming workplace essentials. Products like Midjourney and Cluely started for individuals but are now moving into enterprise use. The line between consumer and business AI is disappearing fast.

Key Takeaways

• 60% of AI spend goes to general-purpose (horizontal) tools like OpenAI and Notion.
• Creative and coding AI apps are now used across all teams, not just specialists.
• AI is slowly replacing repetitive jobs with agentic products that do full workflows.
• Vibe coding platforms like Replit show strong enterprise traction.
• Most AI apps start as consumer tools before evolving into enterprise platforms.

What to do

• Focus AI budgets on tools that boost productivity across the company, not just in single roles.
• Watch emerging “agentic” startups that automate full workflows - they’ll shape the next wave of software.
• Explore creative and coding AI platforms that enable any employee to build or design.
• Expect fast-moving consumer apps to become valuable enterprise tools.
• Stay ready to “hire” AI agents for legal, HR, or customer service tasks.

‍