Vercel Competes With Coding Agents

This reveals that Vercel wins more by becoming the default place agents ship code than by winning prompt to app generation alone. In practice, app builders like v0, Bolt.new, and Lovable are often the first step, but the heavier ongoing work moves into coding agents that edit real repos, run tests, and redeploy repeatedly. That is why Claude Code drove 75% of agent initiated deployments on Vercel, while Lovable and v0 together accounted for 6%.1 2 6

• App builders and coding agents solve different moments in the workflow. v0 is strongest when a team wants a React or Next.js interface fast. Coding agents matter more once the app needs repeated edits across many files, debugging, database changes, and ongoing maintenance inside an existing codebase.2 3 4 6
• That split changes where Vercel competes. Against Lovable or Bolt.new, the fight is over who creates the first working app. Against Claude Code, Cursor, and Replit Agent, the fight is over which tool keeps touching the code every day, because that tool drives far more deployments and infrastructure spend.1 2 5 6
• Vercel has adapted by turning rival app builders into feeders for its hosting layer. Lovable added deployment to Vercel through GitHub sync in July 2026, which means Vercel can still capture deployment revenue even when the app was not originally built in v0.1 7

Going forward, the center of gravity will keep shifting from one shot app generation to agent run software maintenance. That favors platforms like Vercel that sit in the deploy loop for many tools at once, but it also means the most important competitors will be the coding agents that decide where code gets written, tested, and shipped every day.1 2 5 7