Training Resolutions by Model Type Training resolution affects model accuracy, inference speed, and training time. Each model architecture has a default resolution that balances these factors. By default, Roboflow suggests the default training resolution for the selected model architecture.
The table below shows the default training resolution for each model architecture and size. You can override these defaults by configuring the resize preprocessing step when creating a new Dataset Version .
Object Detection Model Type Family & Size Default Training Resolution Object Detection RF-DETR Nano 384x384 Object Detection RF-DETR Small 512x512 Object Detection RF-DETR Medium 576x576 Object Detection RF-DETR Large 704x704 Object Detection RF-DETR X Large 700x700 Object Detection RF-DETR 2X Large 880x880 Object Detection Roboflow 3.0 - Fast 640x640 Object Detection Roboflow 3.0 - Accurate 640x640 Object Detection Roboflow 3.0 - Medium 640x640 Object Detection Roboflow 3.0 - Large 640x640 Object Detection Roboflow 3.0 - Extra Large 640x640 Object Detection YOLOv26 (n/s/m/l/x) 640x640 Object Detection YOLOv12 (n/s/m/l/x) 640x640 Object Detection YOLOv11 (n/s/m/l/x) 640x640 Object Detection YOLOv10 (n/s/m/b/l/x) 640x640 Object Detection YOLOv9 (s/m/c/e) 640x640 Object Detection YOLOv8 (n/s/m/l/x) 640x640 Object Detection YOLOv5 (n/s/m/l/x) 640x640 Object Detection YOLOv7 (legacy) 640x640 Object Detection YOLO-NAS Small 640x640 Object Detection YOLO-NAS Medium 640x640 Object Detection Roboflow Instant 1008x1008
Instance Segmentation Model Type Family & Size Default Training Resolution Instance Segmentation RF-DETR Seg Nano 312x312 Instance Segmentation RF-DETR Seg Small 384x384 Instance Segmentation RF-DETR Seg Medium 432x432 Instance Segmentation RF-DETR Seg Large 504x504 Instance Segmentation RF-DETR Seg X Large 624x624 Instance Segmentation RF-DETR Seg 2X Large 768x768 Instance Segmentation Roboflow 3.0 - Fast (Seg) 640x640 Instance Segmentation Roboflow 3.0 - Accurate (Seg) 640x640 Instance Segmentation Roboflow 3.0 - Medium (Seg) 640x640 Instance Segmentation Roboflow 3.0 - Large (Seg) 640x640 Instance Segmentation Roboflow 3.0 - Extra Large (Seg) 640x640 Instance Segmentation YOLO-seg (v8/10/11/12) 640x640 Instance Segmentation SAM3 (Segment Anything 3) 1008x1008 Instance Segmentation Semantic segmentation (DeepLabV3+) >= 512x512
Classification & Pose Model Type Family & Size Default Training Resolution Classification & Pose Resnet-18/34/50 224x224 Classification & Pose YOLO-cls (v8/11) 224x224 Classification & Pose Vision Transformer (ViT) 224x224 Classification & Pose YOLO-pose (keypoints) 640x640
Multimodal/VLM Model Type Family & Size Default Training Resolution Multimodal/VLM PaliGemma 2 - 3 B 448x448 Multimodal/VLM PaliGemma 2 - 10 B/28 B 448x448 Multimodal/VLM Florence-2 448x448 Multimodal/VLM QWEN 2.5 VL 448x448 Multimodal/VLM SmolVLM2 384x384