Gemma 3 Instruct: pricing, context window and benchmarks
Gemma 3 Instruct is a multimodal model from Google DeepMind. It offers a 128,000-token context window, costs $3.72 per million input tokens and $14.33 per million output tokens, and averages 74.1 across the index's benchmark suite.
Pricing
Gemma 3 Instruct costs $3.72 per million input tokens and $14.33 per million output tokens. A blended 3:1 input-to-output mix works out to about $6.37 per million tokens. Cached input is listed at $0.35 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $3.72 | per 1M tokens |
| Output | $14.33 | per 1M tokens |
| Cached input | $0.35 | per 1M tokens |
| Blended 3:1 | $6.37 | estimate per 1M tokens |
| 1M in + 1M out | $18.05 | 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 | 75.4 | 321 | 47th | |
| GPQA Diamond | 69.2 | 289 | 52th | |
| SWE-bench Verified | 66.0 | 194 | 68th | |
| HumanEval | 80.5 | 220 | 63th | |
| MATH-500 | 78.9 | 238 | 60th | |
| AIME 2025 | 67.4 | 306 | 49th | |
| LiveCodeBench | 64.4 | 308 | 49th | |
| IFEval | 81.1 | 308 | 49th | |
| TruthfulQA | 75.7 | 314 | 48th | |
| HellaSwag | 84.1 | 277 | 54th | |
| MMMU | 72.4 | 232 | 61th | |
| Arena Elo | 1377.2 | 283 | 53th |
What Gemma 3 Instruct is good for
Gemma 3 Instruct is an open-weight model, so it can be self-hosted or run through a third-party inference provider. That matters for data residency, fine-tuning and long-run cost control. At $3.72 per million input tokens (hosted list price) and a 128,000-token context window, it is a practical option for teams that want more control than a closed API allows.
Strengths
- Large 128,000-token context window for long documents and agent traces.
- Input price of $3.72 per million tokens is at or below the index median.
- Open weights make self-hosting, fine-tuning and data-residency review possible.
- Accepts more than text, including image input where the provider supports it.
Trade-offs
- No standout weakness in the indexed specifications, but real-world behaviour still needs testing.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| DeepSeek-Coder Base | DeepSeek | 32768 | $3.53 | 74.1 |
| GPT-5.1 Nano | OpenAI | 32768 | $9.56 | 74.1 |
| Gemma 3 Instruct | Google DeepMind | 128000 | $3.72 | 74.1 |
| Nova Lite Pro | Amazon | 200000 | $9.69 | 74.1 |
| QwQ Base | Alibaba Cloud | 65536 | $3.68 | 74.1 |
| Reka Flash | Reka AI | 128000 | $9.95 | 74.1 |
FAQ
How much does Gemma 3 Instruct cost?
Gemma 3 Instruct is listed at $3.72 per million input tokens and $14.33 per million output tokens. Cached input is $0.35 per million tokens. A blended 3:1 input-to-output mix is about $6.37 per million tokens.
What is the context window of Gemma 3 Instruct?
Gemma 3 Instruct has a context window of 128,000 tokens and can emit up to 16,384 output tokens per request in this index.
Is Gemma 3 Instruct open weights?
Yes. Gemma 3 Instruct is listed as an open-weight model under a open weights licence.
How does Gemma 3 Instruct score on benchmarks?
Gemma 3 Instruct averages 74.1 across the scored benchmarks in this index, ranking #274 of 600 models. Its strongest relative result is Arena Elo at 1377.2.
What is Gemma 3 Instruct best for?
Gemma 3 Instruct is an open-weight model, so it can be self-hosted or run through a third-party inference provider. That matters for data residency, fine-tuning and long-run cost control. At $3.72 per million input tokens (hosted list price) and a 128,000-token context window, it is a practical option for teams that want more control than a closed API allows.