Schematron V2 Small

Availability: Served on Inference.net's own OpenAI-compatible API at api.inference.net/v1. Same two caveats as the Turbo row, read per SKU rather than assumed. The context window displays as "125K tokens" on both the catalogue and the model page while the record behind them says maxContextSize 128000; those are one number, because the site renders every window as floor(tokens/1024)K (1,000,000 shows as "977K", 200,000 as "195K"), so the published 128,000 is the maker's own figure. The catalogue's releaseDate field says 2026-04-16, which cannot be this SKU's launch — it repeats claude-opus-4-7's date on the same page and predates the SKU's first appearance in the OpenRouter catalogue on 2026-09-12 — so we publish the first provably purchasable date instead. Schematron V1 weights (Schematron-3B, Schematron-8B) are public on HuggingFace; the V2 repo inference-net/schematron-v2-llama-3.2-3b is not in the org's public model list and answers 401, so this row is open_source false.