GPT-5.2 Non-reasoning
| Released | 2025-12-11 |
|---|---|
| Price per 1M | $1.75 in · $14.00 out |
| Overall score | 45.4 (early, few results) |
| Context window | 400,000 tokens |
| Provider | OpenAI |
| Weights | Proprietary |
| Configuration | Non-reasoning |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 45.4 | 10 | |
| Overall: Coding | 45.0 | 9 | |
| Overall: Reasoning | 45.8 | 6 | |
| Overall: Agentic work | 45.0 | 9 | |
| Agents | 45.1 | 4 | |
| Reasoning | 44.7 | 5 | |
| Math | 47.2 | 1 | |
| Value vs price | −9.3 |
Evidence
Agents 45.1 ±4.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| tau2-bench Banking Knowledge | 12.6% | tau2-bench | 2026-05-26 |
| tau2-bench Airline | 52.5% | tau2-bench | 2026-05-26 |
| tau2-bench Telecom | 57.2% | tau2-bench | 2026-05-26 |
| tau2-bench Retail | 75.0% | tau2-bench | 2026-05-26 |
Reasoning 44.7 ±3.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Extended NYT Connections | 14.5% | Lech Mazur | 2026-11-10 |
| ARC-AGI-1 | 12.3% | Epoch AI | 2026-11-11 |
| GPQA Diamond | 73.2% | Epoch AI | 2026-07-13 |
| Chess Puzzles | 2.0% | Epoch AI | 2026-10-13 |
| Mystery Game Puzzles | 14.0% | Epoch AI | 2026-08-27 |
Math 47.2 ±3.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| OTIS Mock AIME 2024-2025 | 62.2% | Epoch AI | 2026-07-13 |
Names matched to this model
gpt-5-2-non-reasoning