MobileCLIP-B

Larger MobileCLIP. Better recall, twice the compute.

MultimodalApache-2.0FP16clipembedding

142K downloads 8.2K deploymentsUpdated Feb 4, 2028

Headline:9.8ms · Apple Neural Engine (M4) · FP16

Overview Benchmarks5 Sim Results Deploy5 Files Discussion23

Cross-chip benchmark matrix

Every supported chip, in matched-pair runs from the Fo’c’sle HIL lab. Sortable by any column — click a header. Cells where the chip can’t run this model show Not supported.

Chip platform	Quant	Latency p50(ms)	Latency p95(ms)	Throughput	Acc. retention(%)	Power(W)	Memory(MB)	Tested by
N NVIDIA Jetson AGX Orin 275 TOPS · Module	FP16	28.7	37.4	35 FPS	99.5	42.2	204	Community
Q Snapdragon 8 Gen 3 NPU 45 TOPS · SoC	FP16	61.5	78.4	16 FPS	99.6	7.4	204	Publisher
Apple Neural Engine (M4) 38 TOPS · SoC	FP16	66.9	84.2	15 FPS	99.9	6.6	204	Publisher
Q Qualcomm QCS8550 48 TOPS · SoC	FP16	67.2	84.3	15 FPS	99.9	13.2	204	Community
N NVIDIA Jetson Orin Nano 40 TOPS · SoM	FP16	93.3	121.0	11 FPS	99.4	14.1	204	Publisher
H Hailo-8 26 TOPS · M.2	Not supported
H Hailo-10H 40 TOPS · M.2	Not supported
Q Qualcomm QCS6490 12 TOPS · SoC	Not supported
N NVIDIA Jetson Thor 2070 TOPS · Module	Not supported
π Raspberry Pi 5 + Hailo HAT 26 TOPS · HAT	Not supported
A Ambarella CV5 16 TOPS · SoC	Not supported
A Ambarella CV72 32 TOPS · SoC	Not supported
M MediaTek Genio 700 4 TOPS · SoC	Not supported
G Google Coral Edge TPU 4 TOPS · USB	Not supported
I Intel Movidius Myriad X 4 TOPS · SoC	Not supported
A AMD Versal AI Edge VE2302 22 TOPS · SoC	Not supported
R Rockchip RK3588 6 TOPS · SoC	Not supported

Leader per columnMS-COCO captions · 224×224 image / 32-token text

HIL conditions

All numbers measured on Fo’c’sle HIL rigs in Tel Aviv (primary), Munich (secondary), and Pittsburgh (robotics). Single-stream, batch-1, real preprocessing, real downstream consumer. p50/p95 are over 10,000-frame steady-state windows after a 30-second warm-up. Power draw is package power, not wall power. Memory footprint is the resident model + activations footprint at peak — not on-disk.

Submitted publisher numbers are accepted only if they reproduce within ±8% of an HIL-lab matched run on the same chip in the same input mode. Otherwise they live separately under the Discussion tab.