Home Model series GPT unsuffixed models GPT-5.2 GPT-5.2 · HealthBench · 48,000+ rubric items across 5,000 multi-turn conversations · Scores and configurations GPT-5.2 · HealthBench · 48,000+ rubric items across 5,000 multi-turn conversations · Scores and configurations GPT-5.2 record for HealthBench · 48,000+ rubric items across 5,000 multi-turn conversations. 2 score records for this release, including 2 exact values (may include historical records). Includes 1 source location.
Scores and evaluation conditions GPT-5.2 · HealthBench scores and configurations Metric and score Configuration and evaluation conditions Sources Length-adjusted score 47.6066% Reported 0.476066 · Unit % Published 2025-12-11
GPT-5.2 Instant (system card) Settings System-card HealthBench table names GPT-5.2 Instant; serving configuration is not stated. Unknowns Exact reasoning effort and serving configuration are not reported. Tools Tool access not recorded Reasoning budget not recorded Framework not recorded Protocol HealthBench, 5,000 potentially multi-turn conversations scored with example-specific rubrics; source decimal fraction converted to percentage. Conditions GPT-5.2 Instant; reasoning effort and serving configuration are not stated. Length-adjusted score 63.3379% Reported 0.633379 · Unit % Published 2025-12-11
GPT-5.2 Thinking (system card) Settings System-card HealthBench table names GPT-5.2 Thinking; serving configuration is not stated. Unknowns Exact reasoning effort and serving configuration are not reported. Tools Tool access not recorded Reasoning budget not recorded Framework not recorded Protocol HealthBench, 5,000 potentially multi-turn conversations scored with example-specific rubrics; source decimal fraction converted to percentage. Conditions GPT-5.2 Thinking; reasoning effort and serving configuration are not stated.
Benchmark versions and metrics Aliases HealthBench
Length-adjusted score · % · Higher is better Raw score · % · Higher is better Mean response length · characters · Ranking direction unconfirmed 长度校正分(0–100) · score · Higher is better 未校正分(0–100) · score · Higher is better 相对 GPT-5.4 的长度校正分差 · points · Ranking direction unconfirmed Unadjusted score · % · Higher is better