API-NBA is the NBA-specific data API from API-Sports (the same company behind API-Football, API-Basketball, and others). Its current major version is v2 at base URL https://v2.nba.api-sports.io/. It exposes leagues, seasons, teams, players, games, statistics, and standings endpoints, covering schedules, scores, team/player data, and regular-season standings. There are two auth paths: connect directly to API-Sports with the `x-apisports-key` header, or go through RapidAPI with `x-rapidapi-key` + `x-rapidapi-host: v2.nba.api-sports.io` — and many who search "api nba" actually land on this API-NBA listing on RapidAPI. Its billing model differs fundamentally from balldontlie: API-NBA meters by requests-per-day and every tier can reach all endpoints. The free tier ($0) gives 100 req/day, Pro ($19/mo) 7,500 req/day, Ultra ($29/mo) 75,000 req/day, and Mega ($39/mo) 150,000 req/day, scaling up to roughly 1.5M req/day. balldontlie, by contrast, meters per minute and gates endpoints by tier. So the selection logic is clear: if you want "all endpoints available, just don’t hammer it," API-NBA’s free tier is better for exploration; if you want high-concurrency live score polling, look at balldontlie’s req/min tiers. Two common gotchas: first, the free 100 req/day is genuinely small — any multi-user or polling app outgrows it right after evaluation, so don’t treat it as a production quota; second, when going through RapidAPI, billing and quotas follow RapidAPI’s system, which doesn’t exactly map to the direct API-Sports tiers — confirm which side meters you before integrating to avoid the "thought it was free but RapidAPI charged me" trap. On data depth, API-NBA is practical for everyday schedules, scores, standings, and team/player stats, but its finest-grained play-by-play and advanced data trail the official stats.nba.com, so heavy analytics may need a supplementary source.