Lynchmark

Last updated

Just another LLM benchmark. These are the categories I value.

When it comes to making changes to large code bases without them breaking.

  • 1. Gemini 2.5 Pro
  • 2. GPT-5

The only LLMs that are even remotely close at successful marketing and natural language when everyone is starting to hate AI speak.

  • 1. Kimi K2

UI Design

  • 1. Claude Sonnet 4.5
  • 2. GPT-5