GPT-5.2: pricing, context window and benchmarks
GPT-5.2 is a frontier model from OpenAI. It offers a 2,000,000-token context window, costs $19.71 per million input tokens and $85.71 per million output tokens, and averages 88.0 across the index's benchmark suite.
Pricing
GPT-5.2 costs $19.71 per million input tokens and $85.71 per million output tokens. A blended 3:1 input-to-output mix works out to about $36.21 per million tokens. Cached input is listed at $1.83 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $19.71 | per 1M tokens |
| Output | $85.71 | per 1M tokens |
| Cached input | $1.83 | per 1M tokens |
| Blended 3:1 | $36.21 | estimate per 1M tokens |
| 1M in + 1M out | $105.42 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 89.9 | 7 | 99th | |
| GPQA Diamond | 84.9 | 6 | 99th | |
| SWE-bench Verified | 80.2 | 5 | 99th | |
| HumanEval | 91.0 | 33 | 95th | |
| MATH-500 | 90.6 | 39 | 94th | |
| AIME 2025 | 90.6 | 4 | 99th | |
| LiveCodeBench | 79.7 | 46 | 92th | |
| IFEval | 93.6 | 16 | 97th | |
| TruthfulQA | 89.7 | 5 | 99th | |
| HellaSwag | 92.6 | 30 | 95th | |
| MMMU | 85.3 | 3 | 100th | |
| Arena Elo | 1478.0 | 23 | 96th |
What GPT-5.2 is good for
GPT-5.2 sits in the frontier tier of the index. It is aimed at work where answer quality matters more than unit cost: hard reasoning, multi-step coding, document analysis and multimodal inputs. At $19.71 per million input tokens and a 2,000,000-token context window, it is a premium option, so reserve it for tasks where a cheaper model measurably fails.
Strengths
- Top-tier average benchmark score (88.0), placing it #6 of 600.
- Large 2,000,000-token context window for long documents and agent traces.
- Accepts more than text, including image input where the provider supports it.
- Reasoning-oriented profile, useful for multi-step problems and structured analysis.
- Coding benchmarks are a relative strength within its tier.
Trade-offs
- Proprietary licence: you cannot download the weights or self-host the model.
- Premium input pricing at $19.71 per million tokens.
- Cost can climb quickly on long outputs and high-volume traffic.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Claude Opus 4.5 Reasoning | Anthropic | 2000000 | $20.91 | 88.2 |
| GPT-5.2 | OpenAI | 2000000 | $19.71 | 88.0 |
| GPT-5.1 Max | OpenAI | 2000000 | $20.50 | 87.9 |
| Gemini 3 Pro Max | Google DeepMind | 2000000 | $20.57 | 87.9 |
| Claude Opus 4.5 Max | Anthropic | 1000000 | $20.90 | 87.8 |
| GPT-5.1 Pro | OpenAI | 200000 | $19.77 | 87.8 |
FAQ
How much does GPT-5.2 cost?
GPT-5.2 is listed at $19.71 per million input tokens and $85.71 per million output tokens. Cached input is $1.83 per million tokens. A blended 3:1 input-to-output mix is about $36.21 per million tokens.
What is the context window of GPT-5.2?
GPT-5.2 has a context window of 2,000,000 tokens and can emit up to 16,384 output tokens per request in this index.
Is GPT-5.2 open weights?
No. GPT-5.2 is a proprietary model with a proprietary licence, so you cannot download the weights.
How does GPT-5.2 score on benchmarks?
GPT-5.2 averages 88.0 across the scored benchmarks in this index, ranking #6 of 600 models. Its strongest relative result is AIME 2025 at 90.6.
What is GPT-5.2 best for?
GPT-5.2 sits in the frontier tier of the index. It is aimed at work where answer quality matters more than unit cost: hard reasoning, multi-step coding, document analysis and multimodal inputs. At $19.71 per million input tokens and a 2,000,000-token context window, it is a premium option, so reserve it for tasks where a cheaper model measurably fails.