# The Aggregate (theaggregate.ai) > Unified rankings for 1639 ranked models (2135 in the fused corpus) across 23933 public benchmarks, of which 22467 are in the current fit. Refreshed daily. A metric-aware hierarchical continuous-response IRT fit — per-model ability, per-benchmark difficulty, discrimination and noise — maps every model onto one rating scale. Rows below are fused across reasoning-effort settings, one per model, matching the default board on /models. Human reference baselines combine benchmark-specific human results and estimates and are excluded from the ranking. Page HTML served to crawlers contains readable summaries of every model and benchmark. Snapshot: 2026-10-14 ## Top models by Aggregate Rating - Claude Mythos 5.1 (Anthropic): Rating 2018, rank #1 - Claude Mythos 5 (Anthropic): Rating 2004, rank #2 - Claude Opus 5.5 (Anthropic): Rating 1996, rank #3 - GPT-6 Astra Pro (OpenAI): Rating 1985, rank #4 - GPT-6 Astra (OpenAI): Rating 1984, rank #5 - Claude Mythos Preview (Anthropic): Rating 1981, rank #6 - Gemini 4 Argon (Google): Rating 1975, rank #7 - Claude Fable 5.1 (Anthropic): Rating 1964, rank #8 - GPT-6.1 Sol (OpenAI): Rating 1964, rank #9 - Muse Wonder Full (Meta): Rating 1944, rank #10 ## Key pages - [AI subscriptions](https://theaggregate.ai/subscriptions): compare plan prices, usage limits and editorial recommendations - [Model leaderboard](https://theaggregate.ai/models): all rated models - [Benchmarks](https://theaggregate.ai/benchmarks): browse tests by category and find self-run evaluations - [Benchmark catalog](https://theaggregate.ai/benchmarks/catalog): every tracked benchmark with sources and saturation forecasts - [Daily digest](https://theaggregate.ai/whats-new): what changed in the latest scrape - [Methodology](https://theaggregate.ai/how-it-works): how the unified rankings work - [Our benchmarks](https://theaggregate.ai/our-benchmarks): evaluations the site runs itself, including Guesswork - [Sitemap](https://theaggregate.ai/sitemap.xml): every model and benchmark detail page Model pages: https://theaggregate.ai/model?slug= Benchmark pages: https://theaggregate.ai/benchmark?slug=