302.AI
LLM-Stats aggregates upstream OpenAI benchmark data for the base GPT-5.1 model, citing a score of 0.96 on τ²-Bench Telecom and 0.94 on AIME 2025, both attributed to openai.com methodology. The composite LLM Stats Score places GPT-5.1 at rank 90, with quality tracker values near baseline over recent 7-day and 30-day windows. Blended price and performance-by-depth metrics are aggregator-derived and should not be treated as authoritative for any specific provider. The page lists additional benchmark coverage for GPT-5.1, including BrowseComp Long Context at the 128k setting and other capability tiers tracked by the aggregator. Performance across conversation depths shows relatively stable σ values, with chat-related tasks showing the strongest positive deviation. All cited scores trace to official OpenAI sources rather than independent measurements, so they corroborate rather than expand the model family's documented capability profile.
