Gemini 3 Pro: pricing, context window and benchmarks
Gemini 3 Pro is a frontier model from Google DeepMind. It offers a 2,000,000-token context window, costs $18.24 per million input tokens and $82.93 per million output tokens, and averages 86.1 across the index's benchmark suite.
Pricing
Gemini 3 Pro costs $18.24 per million input tokens and $82.93 per million output tokens. A blended 3:1 input-to-output mix works out to about $34.41 per million tokens. Cached input is listed at $2.09 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $18.24 | per 1M tokens |
| Output | $82.93 | per 1M tokens |
| Cached input | $2.09 | per 1M tokens |
| Blended 3:1 | $34.41 | estimate per 1M tokens |
| 1M in + 1M out | $101.17 | 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 | 88.2 | 13 | 98th | |
| GPQA Diamond | 79.0 | 72 | 88th | |
| SWE-bench Verified | 78.5 | 12 | 98th | |
| HumanEval | 91.8 | 26 | 96th | |
| MATH-500 | 89.9 | 49 | 92th | |
| AIME 2025 | 84.3 | 46 | 92th | |
| LiveCodeBench | 83.3 | 13 | 98th | |
| IFEval | 93.7 | 14 | 98th | |
| TruthfulQA | 88.8 | 13 | 98th | |
| HellaSwag | 91.3 | 53 | 91th | |
| MMMU | 78.6 | 82 | 86th | |
| Arena Elo | 1460.7 | 52 | 91th |
What Gemini 3 Pro is good for
Gemini 3 Pro 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 $18.24 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 (86.1), placing it #31 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 $18.24 per million tokens.
- Cost can climb quickly on long outputs and high-volume traffic.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Qwen3-Max Pro | Alibaba Cloud | 400000 | $18.87 | 86.2 |
| o4 Max | OpenAI | 2000000 | $19.51 | 86.2 |
| GPT-5.1 Turbo | OpenAI | 2000000 | $19.19 | 86.1 |
| Gemini 3 Pro | Google DeepMind | 2000000 | $18.24 | 86.1 |
| Claude Sonnet 4.5 | Anthropic | 400000 | $17.84 | 86.0 |
| GPT-5.1 | OpenAI | 2000000 | $19.01 | 86.0 |
FAQ
How much does Gemini 3 Pro cost?
Gemini 3 Pro is listed at $18.24 per million input tokens and $82.93 per million output tokens. Cached input is $2.09 per million tokens. A blended 3:1 input-to-output mix is about $34.41 per million tokens.
What is the context window of Gemini 3 Pro?
Gemini 3 Pro has a context window of 2,000,000 tokens and can emit up to 8,192 output tokens per request in this index.
Is Gemini 3 Pro open weights?
No. Gemini 3 Pro is a proprietary model with a proprietary licence, so you cannot download the weights.
How does Gemini 3 Pro score on benchmarks?
Gemini 3 Pro averages 86.1 across the scored benchmarks in this index, ranking #31 of 600 models. Its strongest relative result is LiveCodeBench at 83.3.
What is Gemini 3 Pro best for?
Gemini 3 Pro 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 $18.24 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.