GoogleReleased Jun 2026Apache 2.0
Gemma 4 12B has no score on the current Vision Evals. On the earlier Vision Evals board (last updated July 9, 2026) it passed 62.7% of visual understanding tasks (#40 of 77).
It is an open-weight model from Google under the Apache 2.0 license, and free for commercial use.
Gemma 4 12B has not yet been evaluated on the current benchmark. The results below are from the legacy version of Vision Evals, our previous benchmark. See the current Vision Evals
| Category | Passed | Score |
|---|---|---|
| Document Understanding | 8 / 9 | 88.9% |
| Object Understanding | 11 / 14 | 78.6% |
| Defect Detection | 11 / 15 | 73.3% |
| Spatial Understanding | 11 / 19 | 57.9% |
| Object Counting | 1 / 10 | 10% |
Scores based on a single evaluation run · Methodology
View all legacy Vision Evals results →Gemma 4 12B is an open-weight multimodal model from Google in the Gemma 4 family. It is intended for text and image understanding tasks such as visual question answering, OCR, captioning, and document understanding, with a smaller parameter footprint than the larger Gemma 4 variants.
Gemma 4 12B is released under the Apache 2.0 license, so it's commercially usable out of the box: you can ship it in closed-source, commercial products, including self-hosted deployments, without paying a license fee or open-sourcing your code. The Gemma 4 12B license is permissive, so no separate commercial license is required.
Apache-2.0 grants an express patent license that terminates if you bring a patent claim over the work, and it disclaims warranties. Validate Gemma 4 12B on your own data before you depend on it in production.
Read the full Apache 2.0 license ↗This is the straightforward case: a permissive license is the best technical solution and you are free to deploy Gemma 4 12B commercially without open-sourcing your own code.
Do not hesitate to reach out with questions for your commercial project. Our team will help you start solving business problems on the first call. See Roboflow commercial licensing for the models included in each plan.
Talk to salesThis model is released under the Apache License 2.0, a permissive open-source license that allows commercial use, modification, distribution, and patent use.
Yes. Under the terms of the Apache 2.0 license, you can freely use this model for commercial purposes, including in proprietary products. You must retain the copyright notice and disclaimers when redistributing.
License information is provided as a guide and is not legal advice.
Estimated cost per task vs. Visual Understanding score, for this model and others ranked near it. Upper-left is the sweet spot (high quality, low cost). Based on Vision Evals (legacy) results.
10 of 11 models plotted · 1 not yet evaluated
| Model | Score | Median tokens | Est. cost / task | Compare |
|---|---|---|---|---|
| Claude Opus 4.8 | 67.2% | 2.2K | $0.012 | Compare |
| Claude Opus 4.7 | 67.2% | 2.6K | $0.015 | Compare |
| Gemma 4 31B | 67.2% | 467 | $0.0001 | Compare |
| Claude Opus 4.6 | 64.2% | 2.3K | $0.014 | Compare |
| GPT-5.4 Nano | 62.7% | 1.8K | $0.0004 | Compare |
| Gemma 4 12B(this model) | 62.7% | — | — | — |
| Llama 4 Maverick | 59.7% | 2.4K | $0.0004 | Compare |
| Claude Sonnet 4.5 | 59.7% | 2.3K | $0.0092 | Compare |
| Claude Opus 4.1 | 59.7% | 2.1K | $0.040 | Compare |
| Claude Haiku 4.5 | 58.2% | 2.3K | $0.0030 | Compare |
| GPT-5 Nano | 58.2% | 2.7K | $0.0003 | Compare |
Other models worth comparing for similar use cases.