v1.2.0

Added

• Share verified coding tiers across providers by resolving equivalent offerings to one canonical model. The resolver uses exact reviewed Hugging Face checkpoint metadata and reviewed provider aliases; provider-specific rules remove only documented pricing suffixes such as :free or -free, and declared checkpoint metadata takes precedence over name matching. Provider limits and availability remain separate, and missing benchmark evidence stays Unrated.

Full changelog · Compare v1.1.0...v1.2.0

v1.1.0

Added

• Publish provisional coding tiers for resolved models in data/benchmarks/coding.json, including confidence, evidence basis, provider alternatives, and a companion audit file with identity evidence, source revisions, and selection issues.
• Add curated benchmark results and SWE-bench score and tier fields to provider model entries while preserving benchmark values supplied by providers.

Changed

• Move the provider and skipped-provider catalogues into config/. The default provider catalogue is now config/providers.json.

Full changelog · Compare v1.0.0...v1.1.0

v1.0.0

Zero Cost LLMs fetches free model catalogues from 20 providers and publishes one JSON file per provider to the models-data branch each day.

The fetcher uses Python 3.11 and the standard library. It supports provider-specific free-tier checks, optional per-model rate limits, and saved page snapshots for sources that cannot be fetched directly. Each output records provider status and model history; a failed fetch keeps the last successful model list.

Run a fetch for selected providers with --provider. See the README for provider setup and data details.