AI Agent Evaluations

LLM Leaderboard

Performance results of AI coding agents on Sanity development tasks. See how different models perform with Sanity's MCP server and skills.

Our eval suite is open source. Explore the test cases on GitHub.

Trusted by leading digital innovators

  • Figma
  • Anthropic
  • Brex
  • Shopify
  • Linear
  • Skims
  • Unity
  • loveholiday
  • Expedia Group
  • Mejuri
  • Replit
  • Tecovas
  • PUMA
  • MoMA

Ready to build? Let's make it real.

Sign up for Sanity for free and start crafting your first project.

Get started