Memory Usage
infoAll the below benchmarks were performed on iPhone 17 Pro (iOS) and OnePlus 12 (Android).
Classification
| Model | Android (XNNPACK) [MB] | iOS (Core ML) [MB] |
|---|---|---|
| EFFICIENTNET_V2_S | 230 | 87 |
Object Detection
| Model | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| SSDLITE_320_MOBILENET_V3_LARGE | 164 | 132 |
Style Transfer
| Model | Android (XNNPACK) [MB] | iOS (Core ML) [MB] |
|---|---|---|
| STYLE_TRANSFER_CANDY | 1200 | 380 |
| STYLE_TRANSFER_MOSAIC | 1200 | 380 |
| STYLE_TRANSFER_UDNIE | 1200 | 380 |
| STYLE_TRANSFER_RAIN_PRINCESS | 1200 | 380 |
OCR
| Model | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| Detector (CRAFT) + Recognizer (CRNN) | 1400 | 1320 |
Vertical OCR
| Model | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| Detector (CRAFT) + Recognizer (CRNN) | 1000-1600 | 1000-1500 |
LLMs
| Model | Android (XNNPACK) [GB] | iOS (XNNPACK) [GB] |
|---|---|---|
| LLAMA3_2_1B | 3.3 | 3.1 |
| LLAMA3_2_1B_SPINQUANT | 1.9 | 2.4 |
| LLAMA3_2_1B_QLORA | 2.7 | 2.8 |
| LLAMA3_2_3B | 7.1 | 7.3 |
| LLAMA3_2_3B_SPINQUANT | 3.7 | 3.8 |
| LLAMA3_2_3B_QLORA | 3.9 | 4.0 |
Speech to text
| Model | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| WHISPER_TINY | 410 | 375 |
Text to speech
| Model | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| KOKORO_SMALL | 820 | 820 |
| KOKORO_MEDIUM | 1140 | 1100 |
infoThe 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.
Text Embeddings
| Model | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| ALL_MINILM_L6_V2 | 95 | 110 |
| ALL_MPNET_BASE_V2 | 405 | 455 |
| MULTI_QA_MINILM_L6_COS_V1 | 120 | 140 |
| MULTI_QA_MPNET_BASE_DOT_V1 | 435 | 455 |
| CLIP_VIT_BASE_PATCH32_TEXT | 200 | 280 |
Image Embeddings
| Model | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| CLIP_VIT_BASE_PATCH32_IMAGE | 345 | 340 |
Image Segmentation
warningData 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 | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| DEELABV3_RESNET50 | 930 | 660 |
Text to image
| Model | Android (XNNPACK) [MB] | iOS (XNNPACK) [MB] |
|---|---|---|
| BK_SDM_TINY_VPRED_256 | 2400 | 2400 |
| BK_SDM_TINY_VPRED | 6210 | 6050 |