Gemini 3.5 Flash vs Gemma 4 31B
Compare Gemini 3.5 Flash and Gemma 4 31B side-by-side. See how these vision models stack up in Open Prompt, Image Captioning, OCR, Classification, and Object Detection.
Compare Gemini 3.5 Flash vs Gemma 4 31B live
Run the same image across every model that supports a task and compare their outputs side-by-side.
Detect and compare bounding boxes across models on the same image.
Upload an image
Drag and drop an image here, or click to browse
Models in this comparison
Gemini 3.5 Flash vs Gemma 4 31B on Vision Evals
Gemini 3.5 Flash scores higher on 5 of the six Vision Evals tasks.
The widest gap is Reasoning, where Gemini 3.5 Flash leads 82.1% to 50.8%.
Overall, Gemini 3.5 Flash averages 86.0% (#2 of 53) against 67.0% (#30 of 53) for Gemma 4 31B.
Gemma 4 31B is cheaper ($0.0012 vs $0.011 per sample), while Gemini 3.5 Flash is faster (14.8s vs 28.8s per sample).
Gemini 3.5 Flash vs Gemma 4 31B Comparison Table
Evals updated September 5, 2026Pricing updated September 19, 2026
| Property | Gemini 3.5 Flash | Gemma 4 31B |
|---|---|---|
| Organization | ||
| Category | closed | open |
| Modality | multimodal | multimodal |
| Release Date | May 2026 | Apr 2026 |
| Context Window | 1.0M | 256K |
| Parameters | 31B | |
| License | Proprietary | Apache 2.0 |
| Pricing per 1M tokens | ||
| Input $/1M | $1.50 | $0.090 |
| Output $/1M | $9.00 | $0.340 |
| Vision Tasks | ||
| Captioning | Demo | Demo |
| Chart Question Answering | ||
| Classification | Demo | Demo |
| Document Question Answering | ||
| Image Tagging | ||
| Multi-Label Classification | ||
| Object Detection | Demo | Demo |
| OCR | Demo | Demo |
| Vision Language | ||
| Visual Question Answering | Demo | Demo |
| Model Features | ||
| Foundation Vision | ||
| LLMs with Vision Capabilities | ||
| Multimodal Vision | ||
Vision Evalsground-truth scores across 6 vision tasks, pooled at low effort | ||
| Overall | 86.0% | 67.0% |
| Quantizationsself-hosted | ||
| Avg cost / sample | $0.011 | $0.0012 |
| Avg speed / sample | 14.77s | 28.79s |
| By task | ||
| Object Detection (low) | 70.6% ±2.0, Mean of 3 runs, range 68.7 to 72.6 | 48.2% ±0.2, Mean of 3 runs, range 48.0 to 48.4 |
| Object Detection (high) | 69.8% ±1.8, Mean of 3 runs, range 67.5 to 71.1 | – |
| Counting (low) | 80.6% ±0.7, Mean of 3 runs, range 79.7 to 81.1 | 51.4% ±1.4, Mean of 3 runs, range 50.0 to 52.7 |
| Counting (high) | 82.4% ±0.0, Mean of 3 runs, range 82.4 to 82.4 | – |
| Identification (low) | 99.0% ±1.6, Mean of 3 runs, range 96.9 to 100.0 | 80.2% ±3.1, Mean of 3 runs, range 78.1 to 84.4 |
| Identification (high) | 97.9% ±1.6, Mean of 3 runs, range 96.9 to 100.0 | – |
| OCR (low) | 89.3% ±1.6, Mean of 3 runs, range 88.0 to 91.1 | 90.8% ±0.2, Mean of 3 runs, range 90.6 to 90.9 |
| OCR (high) | 88.9% ±0.2, Mean of 3 runs, range 88.7 to 89.1 | – |
| Data Extraction (low) | 94.5% ±0.5, Mean of 3 runs, range 93.8 to 94.8 | 80.4% ±2.6, Mean of 3 runs, range 77.3 to 82.5 |
| Data Extraction (high) | 95.5% ±1.5, Mean of 3 runs, range 93.8 to 96.9 | – |
| Reasoning (low) | 82.1% ±2.0, Mean of 3 runs, range 80.1 to 84.1 | 50.8% ±1.7, Mean of 3 runs, range 49.0 to 52.3 |
| Reasoning (high) | 81.0% ±1.7, Mean of 3 runs, range 79.5 to 82.8 | – |
Gemini 3.5 Flash vs Gemma 4 31B: Overview
Gemini 3.5 Flash is a multimodal language model developed by Google DeepMind and released at Google I/O 2026. It is built on the Gemini 3 Flash reasoning foundation and introduces configurable thinking levels (minimal, low, medium, and high) that allow developers to tune the depth of internal reasoning before a response is generated. The model accepts text, image, video, audio, and PDF inputs and produces text output, with a 1 million token context window and up to 65,000 output tokens per request. It is natively multimodal, processing visual inputs alongside text to support tasks such as image captioning, classification, optical character recognition, object detection, and visual grounding, where the model references specific regions within an image or video frame.
Its vision capabilities extend to interpreting UI screenshots, diagrams, charts, and real-world scenes, as well as understanding video and live frame sequences for activity and scene recognition. The model supports combined tool use, including Google Search, URL context, code execution, and custom functions, within a single request, and it uses reasoning context from previous turns when thought signatures are present in the conversation history, enabling persistent multi-turn reasoning chains. Gemini 3.5 Flash carries a knowledge cutoff of January 2026 and is available via the Gemini API, Google AI Studio, Google Antigravity, and the Gemini Enterprise Agent Platform.
Gemma 4 31B is the largest dense model in Google's Gemma 4 family, built from the same research as Gemini 3 and released as open weights under the Apache 2.0 license. It supports a 256K token context window with text and image input, configurable thinking mode for step-by-step reasoning, and multilingual support across 140+ languages. The unquantized model fits on a single 80GB GPU.
For vision tasks, Gemma 4 31B supports image understanding with variable aspect ratios and resolutions, and can output structured bounding boxes for UI element detection, making it useful for document parsing and UI understanding. Compared to Gemma 3, it delivers stronger reasoning and multimodal performance. It is part of a four-size family alongside the 26B A4B MoE variant and two on-device models (E2B, E4B), with the 31B dense variant optimized for output quality and fine-tuning over inference speed.
Frequently Asked Questions
On Roboflow's Vision Evals, Gemini 3.5 Flash performed better. It scores higher on 5 of the six vision tasks and averages 86.0% (#2 of 53) against 67.0% (#30 of 53) for Gemma 4 31B. These scores measure vision capabilities only; either model may perform differently on work outside this benchmark's tasks.
Yes. On the Vision Evals Reasoning benchmark at low effort, Gemini 3.5 Flash leads with 82.1% against 50.8%. This is the widest gap between the two models across the benchmark's tasks.
Gemma 4 31B is cheaper to run. Across Roboflow's Vision Evals task mix it averaged $0.0012 per sample against $0.011. Actual costs depend on your image sizes, prompts, and output length.
Gemini 3.5 Flash is faster. Across Roboflow's Vision Evals it averaged 14.8s per inference against 28.8s. Latency varies with image size, prompt length, and provider load, so treat these as relative rather than guaranteed figures.
Yes. The comparison demo on this page runs both models on the same image side by side for open prompts and image captioning in the free Roboflow Playground. You can try it instantly, and a free account unlocks unlimited runs.