LLM Reference

LLM Reference is a reference site for people who need to compare large language models without sifting through press releases and fragmented docs. It collects model specs, benchmark scores and pricing updates so you can see how options stack up for tasks like coding, long-context work, vision and image or video generation.

The site keeps a running catalogue of models and providers and updates benchmark results and price changes as they appear. The listing notes coverage across many models and providers, and organises results into searchable leaderboards and category pages to make side-by-side comparisons easier.

• Model directory and side‑by‑side comparisons by capability or use case
• Continuously refreshed benchmark scores from multiple evaluation suites
• Live pricing and provider cost comparisons
• Editorial picks and curated recommendations for common tasks

LLM Reference aims to be useful for developers, product teams, researchers and creative professionals who need a single place to check performance and cost tradeoffs. The official site is https://www.llmreference.com/.