Take the data. It is yours.

One object per tool under tools. Prices are US dollars per month before add-ons and annual discounts; 0 is a real $0 plan or free software, null means the vendor publishes no flat monthly price. The CSV carries the same column names.

30 fields per record in tools: slug, name, category, official_site, helpcompare_url, entry_price_usd, entry_price_unit, entry_price_unit_label, entry_price_included_users, entry_price_min_users, entry_price_unit_source, entry_price_text, entry_price_source, free_plan, free_trial, pricing_model, ai_feature, ai_pricing_model, ai_pricing_detail, rating, review_count, rating_source, rating_listing, rating_url, rating_checked, rating_note, last_verified, comparisons, migrations_from, migrations_to — definitions (JSON Schema)

One object per AI feature under records. pricing_model_key is one of the four billing models (Bundled free, Flat-fee add-on, Metered, Custom quote); detail holds the plain-English invoice line.

8 fields per record in records: tool, category, ai_feature, pricing_model, pricing_model_key, detail, tool_url, last_verified — definitions (JSON Schema)

Two arrays: comparisons (below) and migrations (slug, from, to, url). Together they are the edge list of the software graph.

5 fields per record in comparisons: slug, title, url, last_verified, tools — definitions (JSON Schema)

A JSON array (no wrapper object), one snapshot per row: the cheapest published entry price in a category on the day it was captured. price is a decimal string in US dollars; unit is the billing unit. GET only, no key, no rate limit.

5 fields per row: captured_date, category, tool, price, unit — definitions (JSON Schema)