My LLM Tier List

2026-02-05

Summary: A sourced summary of Daniil Sedov's note about keeping a personal, time-indexed ranking of frontier AI models based on real usage.

This sourced note explains why a static recommendation list for AI models becomes outdated quickly. The original author created a living LLM tier list because new frontier models are released often, and each model can feel different depending on the task. The main lesson for readers is simple: model rankings are subjective. They depend on what you need most, such as instruction following, coding, reasoning, agentic work, speed, cost, or taste.

Free Basics version: use this post as a reminder to test AI tools on your own real tasks instead of trusting one fixed leaderboard forever.

Source: Daniil Sedov, Gusarich's thoughts, gusarich.com/blog/my-llm-tier-list