o3: pricing, context window and benchmarks
o3 is a frontier model from OpenAI. It offers a 400,000-token context window, costs $16.80 per million input tokens and $81.80 per million output tokens, and averages 84.5 across the index's benchmark suite.
Pricing
o3 costs $16.80 per million input tokens and $81.80 per million output tokens. A blended 3:1 input-to-output mix works out to about $33.05 per million tokens. Cached input is listed at $2.22 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $16.80 | per 1M tokens |
| Output | $81.80 | per 1M tokens |
| Cached input | $2.22 | per 1M tokens |
| Blended 3:1 | $33.05 | estimate per 1M tokens |
| 1M in + 1M out | $98.60 | 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 | 86.9 | 39 | 94th | |
| GPQA Diamond | 78.3 | 80 | 87th | |
| SWE-bench Verified | 72.9 | 64 | 89th | |
| HumanEval | 89.8 | 51 | 92th | |
| MATH-500 | 87.9 | 69 | 89th | |
| AIME 2025 | 84.4 | 45 | 93th | |
| LiveCodeBench | 80.7 | 35 | 94th | |
| IFEval | 88.7 | 90 | 85th | |
| TruthfulQA | 84.6 | 72 | 88th | |
| HellaSwag | 92.7 | 28 | 95th | |
| MMMU | 82.3 | 30 | 95th | |
| Arena Elo | 1449.0 | 75 | 88th |
What o3 is good for
o3 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 $16.80 per million input tokens and a 400,000-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 (84.5), placing it #53 of 600.
- Large 400,000-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 $16.80 per million tokens.
- Cost can climb quickly on long outputs and high-volume traffic.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Gemini 3 Flash Pro | Google DeepMind | 400000 | $17.71 | 84.8 |
| GPT-4.1 Max | OpenAI | 400000 | $17.34 | 84.7 |
| Qwen3-Max Turbo | Alibaba Cloud | 1000000 | $17.70 | 84.7 |
| o3 | OpenAI | 400000 | $16.80 | 84.5 |
| o4 Vision | OpenAI | 200000 | $17.01 | 84.5 |
| Qwen3-Max | Alibaba Cloud | 400000 | $17.37 | 84.4 |
FAQ
How much does o3 cost?
o3 is listed at $16.80 per million input tokens and $81.80 per million output tokens. Cached input is $2.22 per million tokens. A blended 3:1 input-to-output mix is about $33.05 per million tokens.
What is the context window of o3?
o3 has a context window of 400,000 tokens and can emit up to 4,096 output tokens per request in this index.
Is o3 open weights?
No. o3 is a proprietary model with a proprietary licence, so you cannot download the weights.
How does o3 score on benchmarks?
o3 averages 84.5 across the scored benchmarks in this index, ranking #53 of 600 models. Its strongest relative result is Arena Elo at 1449.0.
What is o3 best for?
o3 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 $16.80 per million input tokens and a 400,000-token context window, it is a premium option, so reserve it for tasks where a cheaper model measurably fails.