FocsleFocsle
UL

YOLO26-Nano

6%
by Ultralytics

Tiny variant tuned for sub-watt accelerators. Where the Coral TPU still earns its place.

Object detectionMITINT8INT4cocotinylow-power
1.2M downloads 71K deploymentsUpdated Mar 22, 2028
Headline:4.8ms · Google Coral Edge TPU · 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
Q
Qualcomm QCS8550
48 TOPS · SoC
INT84.716.01212 FPS98.512.94Community
A
Ambarella CV72
32 TOPS · SoC
INT86.257.98160 FPS97.75.24Community
π
Raspberry Pi 5 + Hailo HAT
26 TOPS · HAT
INT87.709.62130 FPS97.96.94Publisher
H
Hailo-8
26 TOPS · M.2
INT88.2810.5121 FPS97.52.34Publisher
Q
Qualcomm QCS6490
12 TOPS · SoC
INT812.515.280 FPS98.98.24Fo’c’sle HIL
A
Ambarella CV5
16 TOPS · SoC
INT814.318.270 FPS97.24.04Community
R
Rockchip RK3588
6 TOPS · SoC
INT819.825.951 FPS98.09.34Publisher
G
Google Coral Edge TPU
4 TOPS · USB
INT824.130.142 FPS97.81.54Community
I
Intel Movidius Myriad X
4 TOPS · SoC
INT826.332.338 FPS97.82.44Fo’c’sle HIL
M
MediaTek Genio 700
4 TOPS · SoC
In progress
H
Hailo-10H
40 TOPS · M.2
Not supported
Q
Snapdragon 8 Gen 3 NPU
45 TOPS · SoC
Not supported
N
NVIDIA Jetson Orin Nano
40 TOPS · SoM
Not supported
N
NVIDIA Jetson AGX Orin
275 TOPS · Module
Not supported
N
NVIDIA Jetson Thor
2070 TOPS · Module
Not supported
A
AMD Versal AI Edge VE2302
22 TOPS · SoC
Not supported
Apple Neural Engine (M4)
38 TOPS · SoC
Not supported
Leader per columnCOCO 2017 val · 1080p input · single-stream, batch=1

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.