Models Overview

The inference-models library supports a wide range of computer vision models across multiple tasks and backends.

Supported Tasks

• Object Detection: Detect and localize objects in images
• Instance Segmentation: Detect objects with pixel-level masks
• Semantic Segmentation: Classify every pixel in an image
• Classification: Classify entire images or image regions
• Embeddings: Generate vector representations for images and text
• OCR & Document Parsing: Extract text and structure from documents
• Interactive Segmentation: Interactive and automatic segmentation
• Vision-Language Models: Multi-modal understanding and generation
• Depth Estimation: Predict depth maps from images
• Specialized: Gaze detection, face detection, and more

Model Catalog

Legend: ✅ Available | ❌ Not available | 🔑 Requires API key | 📤 Upload only

Object Detection

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
RF-DETR	torch, onnx, trt	Apache 2.0	N/A	✅	✅
YOLOv5	onnx	AGPL-3.0	✅	❌	📤
YOLOv8	onnx, torch-script, trt	AGPL-3.0	✅	✅	✅
YOLOv9	onnx, torch-script, trt	GPL-3.0	❌	❌	📤
YOLOv10	onnx, trt	AGPL-3.0	✅	✅	📤
YOLOv11	onnx, torch-script, trt	AGPL-3.0	✅	✅	✅
YOLOv12	onnx, torch-script, trt	AGPL-3.0	✅	❌	✅
YOLO-NAS	onnx, trt	Apache 2.0	N/A	✅	✅
Grounding DINO	torch	Apache 2.0	N/A	🔑	❌
OWLv2	hugging-face	Apache 2.0	N/A	🔑	❌
Roboflow Instant	hugging-face	Roboflow	✅	❌	✅

Instance Segmentation

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
RF-DETR Seg	torch	Apache 2.0	N/A	✅	✅
YOLOv5 Seg	onnx	AGPL-3.0	✅	❌	✅
YOLOv7 Seg	onnx	AGPL-3.0	❌	❌	✅
YOLOv8 Seg	onnx, torch-script, trt	AGPL-3.0	✅	✅	✅
YOLOv11 Seg	onnx, torch-script, trt	AGPL-3.0	✅	✅	✅
YOLACT	onnx	MIT	N/A	❌	✅

Classification

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
ResNet	torch, onnx, trt	Apache 2.0	N/A	✅	✅
ViT	torch	Apache 2.0	N/A	✅	✅
DINOv3	torch	Meta DINO	N/A	❌	✅
YOLOv8 Cls	onnx, trt	AGPL-3.0	✅	✅	✅

Embeddings

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
CLIP	torch, onnx	MIT	N/A	✅	❌
Perception Encoder	torch	FAIR Noncommercial	❌	✅	❌

Semantic Segmentation

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
DeepLabV3+	torch, onnx, trt	MIT	N/A	❌	✅

OCR & Document Parsing

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
DocTR	torch	Apache 2.0	N/A	🔑	❌
EasyOCR	torch	Apache 2.0	N/A	🔑	❌
TrOCR	hugging-face	MIT	N/A	🔑	❌

Interactive Segmentation

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
SAM	torch	Apache 2.0	N/A	🔑	❌
SAM2	torch	Apache 2.0	N/A	🔑	❌
SAM2 RT	torch	Apache 2.0	N/A	✅	❌

Vision-Language Models

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
Florence-2	torch	MIT	N/A	✅	✅
PaliGemma	torch	Gemma License	N/A	✅	✅
Qwen2.5-VL	torch	Apache 2.0	N/A	✅	✅
Qwen3-VL	torch	Apache 2.0	N/A	✅	✅
Qwen3.5	torch	Apache 2.0	N/A	✅	✅
SmolVLM	torch	Apache 2.0	N/A	✅	✅
Moondream2	torch	Apache 2.0	N/A	✅	❌

Depth Estimation

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
Depth Anything V2	torch	Apache 2.0	N/A	🔑	❌
Depth Anything V3	torch	Apache 2.0	N/A	🔑	❌

Specialized Models

Model	Backends	License	Commercial License in RF Plan	Pre-trained Weights	Trainable at RF
L2CS	torch	MIT	N/A	✅	❌
MediaPipe Face	mediapipe	Apache 2.0	N/A	✅	❌

Next Steps

• Browse individual model pages for detailed documentation
• See Quick Overview for usage examples
• Check Installation Guide for backend setup