GPT-5.5: benchmark results
OpenAI's standard GPT-5.5 model for general-purpose reasoning, coding, and writing. Provider: OpenAI. Released 2026-04-23. Access: API.
Unified ELO 1743 ± 1, rank #17 of 1392 rated models, from 863 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AGC-Bench - arastories | 1.28 | Dataset z-score | 100 |
| AGC-Bench - brainteaser | 1.84 | Dataset z-score | 100 |
| AGC-Bench - liveideabench | 1.57 | Dataset z-score | 100 |
| AGC-Bench - newyorker_humor | 1.06 | Dataset z-score | 100 |
| AGC-Bench - showerthoughts | 1.6 | Dataset z-score | 100 |
| AI for Education Pedagogy | 92.1 | Accuracy (%) | 100 |
| AI for Education Pedagogy - Primary | 96.71 | Accuracy (%) | 100 |
| AI for Education Visual Maths - Statistics and Probability | 85.71 | Accuracy (%) | 100 |
| ActiveVision | 10.6 | Accuracy (%) | 100 |
| Age of LLM | 3 | Points per match (self-reported) | 100 |
| AtmosCoder-Bench | 97.6 | Accuracy (%, mean of 3 runs) | 100 |
| AutomataBench | 45.28 | Weighted pass@1 (%) | 100 |
Interactive version: theaggregate.ai/model?slug=gpt-5-5 · How It Works · Data refreshed daily, snapshot 2026-09-05.