OpenAI·released 2025-12-111 source
GPT-5.2 benchmark scores: 26 benchmarks tracked, leading 1. 23% of its scores are independently reproduced — the rest are self-reported or unverified.
Leads on 1 of 26 benchmarks·best result 96.11 on OTIS Mock AIME 2024-2025 (reproduced)·6 of 26 independently reproduced·$1.75/$14 per M tokens
Head-to-headGPT-5.2 vs Claude Opus 4.5Consensus: LiteLLM · Cross-check: models.dev, OpenRouter · See every model’s pricing →
Leads on: GDPval
| Benchmark | Score | Evidence status | Measured |
|---|---|---|---|
| OTIS Mock AIME 2024-2025 | 96.11accuracy (%) | reproduced· optimizedT1 | 2025-12-11Epoch AI |
| GPQA diamond | 88.53accuracy (%) | reproduced· optimizedT1 | 2025-12-11Epoch AI |
| ARC-AGI | 86.2accuracy (%) | unverified· optimizedT1 | 2025-12-11ARC Prize Leaderboard |
| DTBench | 84.88accuracy (%) | unverifiedT1 | 2025-12-11https://conceptualreasoning.ai/dtbench |
| VPCT | 76accuracy (%) | unverifiedT1 | 2025-12-11VPCT leaderboard |
| METR Time Horizons | 75.29accuracy (%) | unverifiedT2 | 2025-12-11 |
| SWE-Bench verified | 73.76accuracy (%) | reproduced· optimizedT1 | 2025-12-11Epoch AI |
| WeirdML | 72.19accuracy (%) | unverifiedT1 | 2025-12-11https://htihle.github.io/weirdml.html |
| FrontierMath-Tiers-1-3-v2-Private | 67.4accuracy (%) | reproducedT1 | 2025-12-11Epoch AI |
| Terminal Bench | 64.94accuracy (%) | unverifiedT1 | 2025-12-11https://www.tbench.ai/leaderboard/terminal-bench/2.0 |
| ARC-AGI-2 | 52.91accuracy (%) | unverifiedT2 | 2025-12-11 |
| LMCA | 51.66accuracy (%) | unverifiedT1 | 2025-12-11https://conceptualreasoning.ai/lmca |
| GDPval | 49.7accuracy (%) | unverifiedT2 | 2025-12-11 |
| Chess Puzzles | 45.29accuracy (%) | unverifiedT2 | 2025-12-11 |
| DeepResearch Bench | 41.1accuracy (%) | unverifiedT1 | 2025-12-11https://evals.futuresearch.ai/#drb |
| SimpleQA Verified | 37.1accuracy (%) | unverifiedT2 | 2025-12-11 |
| SimpleBench | 34.96accuracy (%) | unverifiedT1 | 2025-12-11SimpleBench Leaderboard |
| FrontierMath-Tier-4-v2-Private | 31.7accuracy (%) | reproducedT1 | 2025-12-11Epoch AI |
| GSO-Bench | 27.45accuracy (%) | unverifiedT1 | 2025-12-11https://gso-bench.github.io/index.html |
| HLE | 24.16accuracy (%) | unverifiedT2 | 2025-12-11 |
| EBR-bench | 23.02accuracy (%) | reproducedT1 | 2025-12-11Epoch AI |
| CL-bench | 18.2accuracy (%) | unverifiedT2 | 2025-12-11 |
| Mystery Game Puzzles | 15.18accuracy (%) | unverifiedT2 | 2025-12-11 |
| ProofBench | 15accuracy (%) | unverifiedT1 | 2025-12-11https://www.vals.ai/benchmarks/proof_bench |
| Furniture Assembly | 11.9accuracy (%) | unverifiedT2 | 2025-12-11 |
| Remote Labor Index | 2.5accuracy (%) | unverifiedT2 | 2025-12-11 |
Claims drawn from cited facts, not live model generation.
This model was developed by OpenAI and released in 2025.
2 cited facts
This model has scores on 26 tracked benchmarks and currently leads 1 of them. The record’s evidence status is independently reproduced.
3 cited facts
Its mean gap to the best recorded score across its tracked benchmarks is 29.9 points.
1 cited fact
6 facts cross-checked across data sources: 2 corroborated, 2 single-source, 2 disputed.
Source facts, citations, and refresh stamp for this record.
Sources: epoch_benchmarks, litellm_prices, modelsdev_models, openrouter_models
GPT-5.2 is an AI model developed by OpenAI, released 2025-12-11. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.
GPT-5.2 has recorded scores on 26 benchmarks, each shown with its evidence status.
GPT-5.2 has recorded scores on 26 benchmarks — GPQA diamond, OTIS Mock AIME 2024-2025, DTBench, WeirdML, LMCA, SimpleQA Verified, and 20 more. The full table above shows each score with its evidence status.
6 of 26 recorded scores (23%) are independently reproduced rather than self-reported by the lab.
GPT-5.2 has 26 tracked claims: 6 independently reproduced, 20 unverified.
Listed API pricing: $1.75 per million input tokens, $14 per million output tokens (prices disputed across sources). See the pricing block for the full breakdown.