FocsleFocsle
HE

SmolVLM-3B

33%
by HuggingFace-Edge

Compact VLM that comfortably runs on a Jetson Orin Nano. The default 'on-device assistant for cameras' pick.

MultimodalApache-2.0INT8MIXEDvlmsmolcaptioning
184K downloads 6.4K deploymentsUpdated Apr 9, 2028
Headline:142ms · NVIDIA Jetson Orin Nano · INT8

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 platformQuantLatency p50(ms)Latency p95(ms)ThroughputAcc. retention(%)Power(W)Memory(MB)Tested by
N
NVIDIA Jetson Thor
2070 TOPS · Module
MIXED110.8131.79 FPS99.3111.04,956Publisher
N
NVIDIA Jetson AGX Orin
275 TOPS · Module
MIXED152.7188.86.6 FPS98.850.54,956Community
H
Hailo-10H
40 TOPS · M.2
INT8315.5403.93.2 FPS98.03.23,540Fo’c’sle HIL
Apple Neural Engine (M4)
38 TOPS · SoC
MIXED342.2421.92.9 FPS99.47.34,956Community
Q
Qualcomm QCS8550
48 TOPS · SoC
INT8348.8451.52.9 FPS98.411.83,540Publisher
Q
Snapdragon 8 Gen 3 NPU
45 TOPS · SoC
INT8352.8434.02.8 FPS97.37.33,540Publisher
N
NVIDIA Jetson Orin Nano
40 TOPS · SoM
INT8422.1535.72.4 FPS98.713.33,540Fo’c’sle HIL
H
Hailo-8
26 TOPS · M.2
Not supported
Q
Qualcomm QCS6490
12 TOPS · SoC
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.