AdvancedIF

reasoning official site →

AdvancedIF is a rubric-based benchmark measuring complex, multi-turn, and system-prompted instruction following ability, scored with a calibrated LLM judge against per-instruction rubrics.

Methodology

Imported from llm-stats public benchmark metadata. Modality: text. Max score: 1. Categories: general, instruction_following, reasoning. Language: en. Verified by llm-stats: no.

Leaderboard

  1. MAI-Thinking-1 self-reported llm-stats
    85.0%
  2. MAI-Code-1-Flash self-reported llm-stats
    71.4%