Gemma 4 31B vs Qwen3.6 35B A3B
Compare Gemma 4 31B and Qwen3.6 35B A3B side-by-side. See how these vision models stack up in Image Captioning, OCR, Open Prompt, Object Detection, and Classification.
Compare Gemma 4 31B vs Qwen3.6 35B A3B 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
Gemma 4 31B vs Qwen3.6 35B A3B on Vision Evals
Qwen3.6 35B A3B scores higher on 5 of the six Vision Evals tasks.
The widest gap is Counting, where Qwen3.6 35B A3B leads 65.3% to 51.4%.
Overall, Gemma 4 31B averages 67.0% (#34 of 59) against 71.9% (#23 of 59) for Qwen3.6 35B A3B.
Qwen3.6 35B A3B is both cheaper ($0.0012 vs $0.0015 per sample) and faster (27.1s vs 34.4s per sample).
Gemma 4 31B vs Qwen3.6 35B A3B Comparison Table
Evals updated September 22, 2026Pricing updated September 23, 2026
| Property | Gemma 4 31B | Qwen3.6 35B A3B |
|---|---|---|
| Organization | Qwen | |
| Category | open | open |
| Modality | multimodal | multimodal |
| Release Date | Apr 2026 | Apr 2026 |
| Context Window | 256K | 262K |
| Parameters | 31B | 35B total, 3B active |
| License | Apache 2.0 | Apache 2.0 |
| Pricing per 1M tokens | ||
| Input $/1M | $0.090 | $0.150 |
| Output $/1M | $0.340 | $1.00 |
| 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 |
| Phrase Grounding | ||
| Video Classification | ||
| Model Features | ||
| Foundation Vision | ||
| LLMs with Vision Capabilities | ||
| Multimodal Vision | ||
Vision Evalsground-truth scores across 6 vision tasks | ||
| Overall | 67.0% | 71.9% |
| Quantizationsself-hosted | ||
| Avg cost / sample | $0.0015 | $0.0012 |
| Avg speed / sample | 34.36s | 27.10s |
| By task | ||
| Object Detection | 47.5% ±0.6, Mean of 3 runs, range 46.8 to 48.0 | 57.0% ±1.3, Mean of 3 runs, range 56.1 to 58.7 |
| Counting | 51.4% ±2.7, Mean of 3 runs, range 48.6 to 54.0 | 65.3% ±2.7, Mean of 3 runs, range 62.2 to 67.6 |
| Identification | 79.2% ±1.6, Mean of 3 runs, range 78.1 to 81.3 | 82.3% ±6.3, Mean of 3 runs, range 75.0 to 87.5 |
| OCR | 90.8% ±0.6, Mean of 3 runs, range 90.2 to 91.5 | 87.7% ±0.0, Mean of 3 runs, range 87.6 to 87.7 |
| Data Extraction | 80.4% ±1.0, Mean of 3 runs, range 79.4 to 81.4 | 84.5% ±1.0, Mean of 3 runs, range 83.5 to 85.6 |
| Reasoning | 52.8% ±1.3, Mean of 3 runs, range 51.7 to 54.3 | 54.8% ±1.3, Mean of 3 runs, range 53.0 to 55.6 |
Gemma 4 31B vs Qwen3.6 35B A3B: Overview
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.
Qwen3.6-35B-A3B is a sparse Mixture-of-Experts (MoE) multimodal language model developed by the Qwen team at Alibaba Group. It carries 35 billion total parameters but activates only approximately 3 billion per forward pass via a learned routing mechanism, giving it the representational capacity of a large dense model at a fraction of the inference compute. The model is natively multimodal, processing images, documents, and video alongside text as a core architectural capability rather than an add-on. It supports a native context window of 262,144 tokens, extensible up to 1,010,000 tokens via YaRN. A key design feature is the unified thinking/non-thinking mode framework: users can switch between deliberate chain-of-thought reasoning and fast direct responses within a single model, and a "thinking preservation" option retains reasoning context across multi-turn agentic workflows to reduce redundant computation.
The model is specifically optimized for agentic coding tasks, including repository-level reasoning, frontend workflow generation, multi-step tool use, and MCP (Model Context Protocol) integration. On SWE-bench Verified it scores 73.4%, on Terminal-Bench 2.0 it scores 51.5%, and on MCPMark it scores 37.0%. For vision-language tasks it achieves 92.0 on RefCOCO, 89.9 on OmniDocBench 1.5, and 83.7 on VideoMMMU. The model also supports Multi-Token Prediction (MTP) for speculative decoding. All Qwen3.6 open-weight models are released under the Apache 2.0 license.