Performance Benchmarks

This page provides a centralized reference for evaluating AI inference performance across MediaTek Genio platforms, which aggregates benchmarking results for various AI workloads, including analytical AI and generative AI, across multiple inference frameworks such as TFLite (LiteRT) and ONNX Runtime.

Important

For more platform-specific details and comprehensive performance data, please refer to the Model Zoo.

AI Supporting Scope

The following table summarizes the AI capabilities and framework support across different MediaTek Genio platforms.

AI Supporting Scope Across Different Platform

Platform	OS	TFLite - Analytical AI (Online)	TFLite - Analytical AI (Offline)	TFLite - Generative AI	ONNX Runtime - Analytical AI
Genio 520/720	Android	CPU + GPU + NPU	NPU	NPU	CPU + NPU
	Yocto	CPU + GPU + NPU	NPU	X (ETA: 2026/Q2)	CPU + NPU
Genio 510/700	Android	CPU + GPU + NPU	NPU	X	X
	Yocto	CPU + GPU + NPU	NPU	X	CPU
	Ubuntu	CPU + GPU + NPU	NPU	X	X
Genio 1200	Android	CPU + GPU + NPU	NPU	X	X
	Yocto	CPU + GPU + NPU	NPU	X	CPU
	Ubuntu	CPU + GPU + NPU	NPU	X	X
Genio 350	Android	CPU + GPU + NPU	X	X	X
	Yocto	CPU + GPU	X	X	CPU
	Ubuntu	CPU + GPU	X	X	X

TFLite(LiteRT) - Analytical AI

The following tables list the validated TFLite analytical models and their performance across Genio platforms. The statistics were measured using offline inference with performance mode enabled.

Models for Detection

Genio 360

Genio 520

Genio 720

Genio 510

Genio 700

Genio 1200

MT8893

Task

Model Name

Source model type

Data Type

Input Size

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Detail

Object Detection

YOLOv5s

.pt

Quant8

640x640

8.68

6.32

5.35

16.00

10.10

19.00

3.42

Link

Object Detection

YOLOv5s

.pt

Float32

640x640

28.25

18.78

16.66

44.90

32.30

37.00

11.40

Link

Object Detection

YOLOv8s

.pt

Quant8

640x640

12.92

9.31

8.04

24.00

17.00

28.40

5.64

Link

Object Detection

YOLOv8s

.pt

Float32

640x640

42.41

27.13

24.41

69.30

50.30

54.60

16.34

Link

Models for Classification

					Genio 360	Genio 520	Genio 720	Genio 510	Genio 700	Genio 1200	MT8893
Task	Model Name	Source model type	Data Type	Input Size	Inference Time (ms)	Inference Time (ms)	Inference Time (ms)	Inference Time (ms)	Inference Time (ms)	Inference Time (ms)	Inference Time (ms)	Detail
Classification	DenseNet	.pt	Quant8	224x224	5.86	4.46	3.83	7.30	5.30	6.00	2.40	Link
Classification	DenseNet	.pt	Float32	224x224	12.07	9.21	7.72	16.30	11.30	12.40	4.74	Link
Classification	EfficientNet	.pt	Quant8	224x224	2.20	1.83	1.56	4.00	3.00	3.00	1.16	Link
Classification	EfficientNet	.pt	Float32	224x224	4.85	3.57	3.15	9.30	6.00	6.10	2.21	Link
Classification	MobileNetV2	.pt	Quant8	224x224	1.28	1.13	0.97	1.00	1.00	1.00	0.78	Link
Classification	MobileNetV2	.pt	Float32	224x224	2.60	2.06	1.71	3.00	2.00	2.10	1.29	Link
Classification	MobileNetV3	.pt	Quant8	224x224	0.91	0.80	0.74	1.00	0.40	0.90	0.64	Link
Classification	MobileNetV3	.pt	Float32	224x224	1.67	1.36	1.20	2.00	1.10	2.00	0.97	Link
Classification	ResNet	.pt	Quant8	224x224	2.25	1.71	1.48	2.00	2.00	2.10	1.08	Link
Classification	ResNet	.pt	Float32	224x224	6.20	4.35	3.86	9.30	6.30	8.40	2.56	Link
Classification	SqueezeNet	.pt	Quant8	224x224	1.55	1.21	1.11	1.00	1.00	1.00	0.86	Link
Classification	SqueezeNet	.pt	Float32	224x224	3.57	2.49	2.31	4.10	3.00	3.00	1.64	Link
Classification	VGG	.pt	Quant8	224x224	17.72	13.47	11.17	N/A	N/A	24.40	6.47	Link
Classification	VGG	.pt	Float32	224x224	69.25	38.44	33.17	80.30	56.30	50.40	19.87	Link

Models for Recognition

Genio 360

Genio 520

Genio 720

Genio 510

Genio 700

Genio 1200

MT8893

Task

Model Name

Source model type

Data Type

Input Size

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Inference Time (ms)

Detail

Recognition

VGGFace

.pt

Quant8

224x224

18.21

13.89

11.50

N/A

N/A

24.40

6.58

Link

Recognition

VGGFace

.pt

Float32

224x224

71.01

39.26

33.80

81.30

56.30

49.50

20.00

Link

TFLite(LiteRT) - Generative AI

For Generative AI workloads, the following tables provide representative performance data for reference and platform capability validation.

LLM Performance Comparison

Prompt Mode Comparison (Unit: tok/s)

Model	Genio 720	Genio 520	MT8893
DeepSeek-R1-Distill-Llama-8B	36.65	Not Support	425.79
DeepSeek-R1-Distill-Qwen-1.5B	341.69	276.82	1057.25
DeepSeek-R1-Distill-Qwen-7B	69.23	Not Support	448.17
gemma2-2b-it	193.39	Not Support	891.00
internlm2-chat-1.8b	276.22	243.61	1544.70
llama3-8b	56.50	Not Support	426.13
llama3.2-1B-Instruct	401.29	335.92	2093.61
llama3.2-3B-Instruct	154.56	Not Support	1022.95
Qwen2-0.5B-Instruct	762.46	641.70	3010.84
Qwen2-1.5B-Instruct	341.99	274.50	1616.22
Qwen2-7B-Instruct	70.42	Not Support	474.38
Qwen1.5-1.8B-Chat	310.64	268.23	1516.50
Qwen2.5-1.5B-Instruct	341.42	269.87	1621.85
Qwen2.5-3B-Instruct	162.48	Not Support	751.06
Qwen2.5-7B-Instruct	70.55	Not Support	471.95
Qwen3 1.7B	233.03	229.63	1069.16
Phi-3-mini-4k-instruct	129.60	Not Support	828.87
MiniCPM-2B-sft-bf16-llama-format	194.79	153.14	886.72
medusa_v1_0_vicuna_7b_v1.5	91.82	Not Support	501.05
vicuna1.5-7b-tree-speculative-decoding-plus	84.90	Not Support	454.58
llava1.5-7b-speculative-decoding	73.10	Not Support	267.98
baichuan-7b-int8-cache	81.18	Not Support	561.76
baichuan-7b	79.75	Not Support	536.64

Generative Mode Comparison (Unit: tok/s)

Model	Genio 720	Genio 520	MT8893
DeepSeek-R1-Distill-Llama-8B	4.58	Not Support	11.36
DeepSeek-R1-Distill-Qwen-1.5B	11.76	9.18	25.68
DeepSeek-R1-Distill-Qwen-7B	4.68	Not Support	11.69
gemma2-2b-it	8.75	Not Support	21.37
internlm2-chat-1.8b	17.55	14.48	42.39
llama3-8b	4.70	Not Support	11.51
llama3.2-1B-Instruct	24.53	20.59	61.14
llama3.2-3B-Instruct	10.58	Not Support	25.05
Qwen2-0.5B-Instruct	50.06	42.43	77.87
Qwen2-1.5B-Instruct	19.56	15.32	38.31
Qwen2-7B-Instruct	4.88	Not Support	11.64
Qwen1.5-1.8B-Chat	9.90	8.63	31.38
Qwen2.5-1.5B-Instruct	18.43	15.11	38.57
Qwen2.5-3B-Instruct	10.31	Not Support	20.87
Qwen2.5-7B-Instruct	4.89	Not Support	11.74
Qwen3 1.7B	10.91	10.53	23.42
Phi-3-mini-4k-instruct	7.32	Not Support	18.87
MiniCPM-2B-sft-bf16-llama-format	7.69	6.48	22.28
medusa_v1_0_vicuna_7b_v1.5	10.56	Not Support	22.79
vicuna1.5-7b-tree-speculative-decoding-plus	12.65	Not Support	22.72
llava1.5-7b-speculative-decoding	7.28	Not Support	6.78
baichuan-7b-int8-cache	4.24	Not Support	11.37
baichuan-7b	4.18	Not Support	10.56

VLM Performance Comparison

ViT Inference Time (Unit: s)

Model	Genio 720	Genio 520	MT8893
Qwen2.5 VL 3B	0.21	Not Support	0.10
InternVL3-1B	1.74	1.94	0.51

Prompt Mode (Unit: tok/s)

Model	Genio 720	Genio 520	MT8893
Qwen2.5 VL 3B	100.07	Not Support	339.90
InternVL3-1B	74.75	65.23	183.64

Generative Mode (Unit: tok/s)

Model	Genio 720	Genio 520	MT8893
Qwen2.5 VL 3B	4.78	Not Support	10.13
InternVL3-1B	6.16	4.52	14.09

Stable Diffusion Performance Comparison

Main Time Comparison (Unit: ms)

Model	Genio 720	Genio 520	MT8893
Stable Diffusion v.1.5	25816	33754	7075
Stable Diffusion v.1.5 controlnet	33642	47923	9395
Stable_diffusion_v1_5_controlnet_lora	34148	Not Support	10268
Stable_diffusion_v1.5_2lora	35978	45619	11487
Stable Diffusion v2.1 base model with controlnet	31183	Not Support	6969
Stable Diffusion v1.5 LCM Ipadaptor	10645	11378	2254
Stable_diffusion_lcm_multiDiffusion	29104	34731	7439

Inference Time Comparison (Unit: ms)

Model	Genio 720	Genio 520	MT8893
Stable Diffusion v.1.5	24813	29001	6132
Stable Diffusion v.1.5 controlnet	32294	37581	8035
Stable_diffusion_v1_5_controlnet_lora	32454	Not Support	8472
Stable_diffusion_v1.5_2lora	33195	39198	10130
Stable Diffusion v2.1 base model with controlnet	29828	Not Support	5451
Stable Diffusion v1.5 LCM Ipadaptor	5861	4978	1077
Stable_diffusion_lcm_multiDiffusion	28127	32585	6698

CLIP Performance Comparison

Main Time Comparison (Unit: ms)

Model	Genio 720	Genio 520	MT8893
img_encoder_proj_clip_vit_large_dynamic	567.61	662.22	358.61
img_encoder_proj_openclip_vit_big_g_dynamic	12035.52	Not Support	1390.56
img_encoder_proj_openclip_vit_h_dynamic	1440.20	Not Support	591.93
text_encoder_clip_vit_large	455.08	748.45	308.72
text_encoder_openclip_vit_h	750.70	Not Support	510.92

Inference Time Comparison (Unit: ms)

Model	Genio 720	Genio 520	MT8893
img_encoder_proj_clip_vit_large_dynamic	257.39	296.84	51.14
img_encoder_proj_openclip_vit_big_g_dynamic	3142.96	Not Support	517.13
img_encoder_proj_openclip_vit_h_dynamic	881.65	Not Support	147.47
text_encoder_clip_vit_large	38.99	45.56	18.94
text_encoder_openclip_vit_h	119.77	Not Support	48.49

ONNX Runtime - Analytical AI

The following tables list ONNX models validated on Genio platforms. Measurements were obtained using the NPU Execution Provider (where available) with performance mode enabled.

Models for TAO Related

Task	Model Name	Data Type	Input Size	G520 NPU (ms)	G520 CPU (ms)	G720 NPU (ms)	G720 CPU (ms)	Pretrained Model
Object Detection	PeopleNet (ResNet34)	Float32	3x544x960	83.51	3460.66	77.63	3422.34	Download
Object Detection	PeopleNet (ResNet34)	Quant8	3x544x960	20.49	911.40	19.44	901.13	Download
Recognition	Action Recognition Net (ResNet18)	Float32	96x224x224	16.83	353.46	14.75	350.78	Download
Pose Estimation	BodyPoseNet	Float32	224x320x3	43.99	1845.36	40.56	1825.80	Download
Object Detection	LPDNet (USA Pruned)	Float32	3x480x640	3.68	107.06	3.42	106.47	Download
Segmentation	PeopleSemSegNet_AMR	Float32	3x576x960	Not Support	7685.97	Not Support	7749.61	Download
Segmentation	PeopleSemSegNet_AMR (Rel)	Float32	3x544x960	15.05	146.83	13.29	136.69	Download
Segmentation	PeopleSemSegNet (ShuffleSeg)	Float32	3x544x960	15.08	140.38	13.41	136.57	Download
Segmentation	PeopleSemSegNet (Vanilla Unet)	Float32	3x544x960	178.51	7510.24	163.45	7346.95	Download
Re-Identification	ReIdentificationNet (ResNet50)	Float32	3x256x128	8.59	237.53	6.87	234.10	Download
OCR	Ocrnet_resnet50	Float32	1x32x100	20.64	300.25	18.16	296.89	Download
OCR	Ocrnet_resnet50 (Pruned)	Float32	1x32x100	14.93	179.51	13.94	175.70	Download
OCR	ocd_resnet50	Float32	3x736x1280	169.41	6520.78	149.85	6340.07	Download
OCR	ocd_resnet50	Float32	3x640x640	76.18	2809.26	68.10	2748.75	Download
OCR	ocdnet_mixnet	Float32	3x640x640	362.87	17742.48	340.09	17436.33	Download
Classification	Pose Classification (ST-GCN)	Float32	3x300x34x1	223.89	787.40	207.00	772.19	Download
Pose Estimation	Centerpose (Chair DLA34)	Float32	3x512x512	Not Support	3035.51	Not Support	2946.96	Download
Pose Estimation	Centerpose (Camera FAN)	Float32	3x512x512	Not Support	7689.45	Not Support	7568.34	Download
Object Detection	LPDNet (CCPD Pruned)	Float32	3x1168x720	7.82	190.42	6.73	186.26	Download
Pose Estimation	Foundation Pose (Refiner)	Float32	6x160x160	64.91	682.93	60.97	674.01	Download
Pose Estimation	Foundation Pose (Score)	Float32	6x160x160	37.37	622.71	34.63	615.33	Download
Pose Estimation	Multi 3D Centerpose	Float32	3x512x512	Not Support	3024.04	Not Support	3005.77	Download

Models for Detection

Task	Model Name	Data Type	Input Size	G520 NPU (ms)	G520 CPU (ms)	G720 NPU (ms)	G720 CPU (ms)	Pretrained Model
Object Detection	YOLOv5s	Quant8	640x640	Not Support	225.74	Not Support	221.29	Download
Object Detection	YOLOv5s	Float32	640x640	36.50	607.68	32.37	586.80	Download
Object Detection	YOLOv8s	Quant8	640x640	90.11	353.19	80.57	346.58	Download
Object Detection	YOLO11s	Quant8	640x640	102.15	301.50	90.99	295.32	Download

Models for Classification

Task	Model Name	Data Type	Input Size	G520 NPU (ms)	G520 CPU (ms)	G720 NPU (ms)	G720 CPU (ms)	Pretrained Model
Classification	ConvNeXt	Quant8	224x224	Not Support	516.21	Not Support	1115.18	Download
Classification	ConvNeXt	Float32	224x224	Not Support	1117.20	Not Support	510.37	Download
Classification	DenseNet	Quant8	224x224	Not Support	104.51	Not Support	103.30	Download
Classification	DenseNet	Float32	224x224	8.46	205.29	7.49	200.32	Download
Classification	EfficientNet	Quant8	224x224	33.33	24.07	30.52	23.94	Download
Classification	EfficientNet	Float32	224x224	3.15	66.64	2.81	65.57	Download
Classification	MobileNetV2	Quant8	224x224	1.43	12.36	1.26	12.23	Download
Classification	MobileNetV2	Float32	224x224	1.75	31.69	1.47	30.41	Download
Classification	MobileNetV3	Quant8	224x224	Not Support	6.30	Not Support	6.16	Download
Classification	MobileNetV3	Float32	224x224	13.72	10.74	12.81	10.45	Download
Classification	ResNet	Quant8	224x224	2.04	45.87	1.78	45.08	Download
Classification	ResNet	Float32	224x224	3.81	112.00	3.49	111.24	Download
Classification	SqueezeNet	Quant8	224x224	9.36	33.08	8.38	31.96	Download
Classification	SqueezeNet	Float32	224x224	9.86	53.15	8.81	52.00	Download
Classification	VGG	Quant8	224x224	13.79	366.54	11.62	366.31	Download
Classification	VGG	Float32	224x224	37.17	902.03	32.24	889.14	Download

Models for Recognition

Task	Model Name	Data Type	Input Size	G520 NPU (ms)	G520 CPU (ms)	G720 NPU (ms)	G720 CPU (ms)	Pretrained Model
Recognition	VGGFace	Quant8	224x224	291.44	366.43	291.22	367.59	Download
Recognition	VGGFace	Float32	224x224	37.98	904.24	32.84	891.02	Download

Models for Robotic

Task	Model Name	Data Type	Input Size	G520 NPU (ms)	G520 CPU (ms)	G720 NPU (ms)	G720 CPU (ms)	Pretrained Model
Omni6DPose	scale_policy	Float32	1x3x3	Not Support	0.18	Not Support	0.17	Download
Diffusion Policy	model_diffusion_sampling	Float32	trajectory:1x16x12, global_cond:1x800	60.66	44.71	57.68	41.96	Download
MobileSam	mobilesam_encoder	Float32	3x448x448	Not Support	705.14	Not Support	694.74	Download
RegionNormalizedGrasp	anchornet	Float32	4x640x360	13.42	187.66	12.13	183.82	Download
RegionNormalizedGrasp	localnet	Float32	64x64x6	20.02	19.65	19.78	20.06	Download
YoloWorld	yoloworld_xl	Float32	3x640x640	465.61	11466.42	403.15	11214.33	Download

Performance Notes

Performance can vary depending on:

• The specific Genio platform and hardware configuration.

• The version of the board image and evaluation kit (EVK).

• The selected backend and model variant.

To obtain the most accurate performance numbers for your use case, you must run the application directly on the target platform.