LLM Tier List

LLM Tier List is a crowd-voted ranking of AI models: browse the leaderboard, cast votes for favorites and least favorites, write reviews, and build shareable tier lists of your own. The current leaderboard already shows real engagement texture, with Claude, GPT variants, Kimi, Qwen, Grok, and DeepSeek models ranked by community votes rather than benchmark scores. The site is explicit about not being affiliated with any AI lab, which matters for a ranking property since vendor-run leaderboards carry obvious bias incentives. What makes this worth watching rather than dismissing as a meme format is that tier lists are how normal people communicate model opinions, far more than MMLU percentages, and aggregated casual opinion is a genuinely different signal from expert evals. The model naming on the leaderboard includes regional Chinese models alongside Western ones, so the voting base looks international. Fun enough to browse, useful as a temperature check on public sentiment between benchmark releases.