LLM Model Database
A maintained reference of frontier and open-weight language models. Pricing, context window, modalities, capabilities, benchmark scores — for every model the major providers ship today. Hit one of the cards to see the per-model page; use the comparison tool for side-by-side analysis.
config/database.php are filled in and the schema imported via
php scripts/install-db.php, the full database renders here.
In the meantime, the underlying data lives in data/seed.sql.
How this database is maintained
Specs and pricing are pulled by hand from each provider's official documentation (platform.openai.com/docs, docs.anthropic.com, ai.google.dev, qwenlm.github.io, and so on). Benchmark scores come from the model card or release announcement on the date of release. Each benchmark result on a model page links to its primary source.
Prices change. The database is refreshed monthly; the "Updated" timestamp on each model page shows the last refresh. Always verify against the provider's own billing page before relying on these numbers for budget commitments.