Memory Usage

info

Memory usage values represent the peak memory increase observed while the model was loaded and actively running inference, relative to the baseline app memory before model initialization.

Classification

Model / Device	iPhone 17 Pro [MB]	Google Pixel 10 [MB]
EFFICIENTNET_V2_S (XNNPACK FP32)	101	122
EFFICIENTNET_V2_S (XNNPACK INT8)	62	78
EFFICIENTNET_V2_S (Core ML FP32)	101	-
EFFICIENTNET_V2_S (Core ML FP16)	87	-

Object Detection

note

Data presented for YOLO models is based on inference with forward_640 method.

Model / Device	iPhone 17 Pro [MB]	Google Pixel 10 [MB]
SSDLITE_320_MOBILENET_V3_LARGE (XNNPACK FP32)	94	104
SSDLITE_320_MOBILENET_V3_LARGE (Core ML FP32)	83	-
SSDLITE_320_MOBILENET_V3_LARGE (Core ML FP16)	62	-
RF_DETR_NANO (XNNPACK FP32)	145	162
YOLO26N (XNNPACK FP32)	36	44
YOLO26S (XNNPACK FP32)	81	82
YOLO26M (XNNPACK FP32)	123	158
YOLO26L (XNNPACK FP32)	170	172
YOLO26X (XNNPACK FP32)	320	309

Style Transfer

Model / Device	iPhone 17 Pro [MB]	Google Pixel 10 [MB]
STYLE_TRANSFER_CANDY (XNNPACK FP32)	1200	1200
STYLE_TRANSFER_CANDY (XNNPACK INT8)	800	800
STYLE_TRANSFER_CANDY (Core ML FP32)	400	-
STYLE_TRANSFER_CANDY (Core ML FP16)	380	-
STYLE_TRANSFER_MOSAIC (XNNPACK FP32)	1200	1200
STYLE_TRANSFER_MOSAIC (XNNPACK INT8)	800	800
STYLE_TRANSFER_MOSAIC (Core ML FP32)	400	-
STYLE_TRANSFER_MOSAIC (Core ML FP16)	380	-
STYLE_TRANSFER_UDNIE (XNNPACK FP32)	1200	1200
STYLE_TRANSFER_UDNIE (XNNPACK INT8)	800	800
STYLE_TRANSFER_UDNIE (Core ML FP32)	400	-
STYLE_TRANSFER_UDNIE (Core ML FP16)	380	-
STYLE_TRANSFER_RAIN_PRINCESS (XNNPACK FP32)	1200	1200
STYLE_TRANSFER_RAIN_PRINCESS (XNNPACK INT8)	800	800
STYLE_TRANSFER_RAIN_PRINCESS (Core ML FP32)	400	-
STYLE_TRANSFER_RAIN_PRINCESS (Core ML FP16)	380	-

OCR

Model / Device	iPhone 17 Pro [MB]	OnePlus 12 [MB]
Detector (CRAFT) + Recognizer (CRNN) (XNNPACK FP32)	1320	1400

Vertical OCR

Model / Device	iPhone 17 Pro [MB]	OnePlus 12 [MB]
Detector (CRAFT) + Recognizer (CRNN) (XNNPACK FP32)	1000-1500	1000-1600

LLMs

Model / Device	iPhone 17 Pro [GB]	OnePlus 12 [GB]
LLAMA3_2_1B (XNNPACK)	3.1	3.3
LLAMA3_2_1B_SPINQUANT (XNNPACK)	2.4	1.9
LLAMA3_2_1B_QLORA (XNNPACK)	2.8	2.7
LLAMA3_2_3B (XNNPACK)	7.3	7.1
LLAMA3_2_3B_SPINQUANT (XNNPACK)	3.8	3.7
LLAMA3_2_3B_QLORA (XNNPACK)	4.0	3.9

Speech to Text

Model / Device	iPhone 17 Pro [MB]	OnePlus 12 [MB]
WHISPER_TINY (XNNPACK)	375	410

Text to Speech

note

The reported memory usage values include the memory footprint of the Phonemis package, which is used for phonemizing input text. Currently, this can range from 100 to 150 MB depending on the device.

Model / Device	iPhone 17 Pro [MB]	OnePlus 12 [MB]
KOKORO_SMALL (XNNPACK)	820	820
KOKORO_MEDIUM (XNNPACK)	1100	1140

Text Embeddings

Model / Device	iPhone 17 Pro [MB]	OnePlus 12 [MB]
ALL_MINILM_L6_V2 (XNNPACK)	110	95
ALL_MPNET_BASE_V2 (XNNPACK)	455	405
MULTI_QA_MINILM_L6_COS_V1 (XNNPACK)	140	120
MULTI_QA_MPNET_BASE_DOT_V1 (XNNPACK)	455	435
CLIP_VIT_BASE_PATCH32_TEXT (XNNPACK)	280	200

Image Embeddings

Model / Device	iPhone 17 Pro [MB]	Google Pixel 10 [MB]
CLIP_VIT_BASE_PATCH32_IMAGE (XNNPACK FP32)	340	345

Semantic Segmentation

note

Data presented in the following sections is based on inference with non-resized output. When resize is enabled, expect higher memory usage and inference time with higher resolutions.

Model / Device	iPhone 17 Pro [MB]	OnePlus 12 [MB]
DEEPLABV3_RESNET50 (XNNPACK)	660	930

Instance Segmentation

note

Data presented in the following sections is based on inference with forward_640 method.

Model / Device	iPhone 17 Pro [MB]	OnePlus 12 [MB]
YOLO26N_SEG (XNNPACK)	668	92
YOLO26S_SEG (XNNPACK)	712	220
YOLO26M_SEG (XNNPACK)	815	570
YOLO26L_SEG (XNNPACK)	1024	680
YOLO26X_SEG (XNNPACK)	1450	1410
RF_DETR_NANO_SEG (XNNPACK)	603	620

Text to Image

Model / Device	iPhone 17 Pro [MB]	OnePlus 12 [MB]
BK_SDM_TINY_VPRED_256 (XNNPACK)	2400	2400
BK_SDM_TINY_VPRED (XNNPACK)	6050	6210