GPT-5.2 (Non-reasoning): benchmark results
GPT-5.2 evaluated with reasoning disabled. Provider: OpenAI. Released 2025-12-11. Access: API.
Aggregate Rating 1629 ± 11, rank #608 of 2168 rated models, from 166 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| CAREBench - Positive Emotion Intensity | 0.21 | Normalized RMSE (0-1; positive emotion intensity, predicted | 100 |
| CAREBench - Positive Emotion Labels | 55.6 | Example-F1 (%; per-narrative F1 of the selected positive emo | 100 |
| CCTU (Non-Thinking) | 16 | Perfect solve rate (%): the share of cases in which the mode | 100 |
| K-Bench (Therapeutic v0) | 98.95 | Overall (%, Therapeutic v0 prompt) | 100 |
| Memora (Standard Inference) - Recommending | 159.28 | Aggregated forgetting-aware memory accuracy (0 to 300) on re | 100 |
| Memora (Standard Inference) - Remembering | 68.63 | Aggregated forgetting-aware memory accuracy (0 to 300) on re | 100 |
| K-Bench - Risk (Therapeutic v0) | 95.15 | Risk Score (%, Therapeutic v0 prompt) | 96.8 |
| K-Bench | 98.59 | Overall (%, Default v1 prompt) | 95.1 |
| ReasonTabQA - English | 61.29 | Answer Accuracy (%; LLM judge against gold answers, program- | 93.8 |
| K-Bench - Risk | 94.71 | Risk Score (%, Default v1 prompt) | 93.4 |
| CAPITU - Easy | 100 | Strict prompt-level accuracy (%) on the 70 easy prompts (one | 85.3 |
| CAPITU - Coherence | 98.4 | Coherence score (%): a GPT-4o-mini judge's 0-1 rating of the | 82.4 |
Interactive version: theaggregate.ai/model?slug=gpt-5-2-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.