Green check
Link copied to clipboard

YOLOv5 Export Competition

We are super excited to announce the first-ever Ultralytics YOLOv5 Export Competition with $10,000.00

Facebook logoTwitter logoLinkedIn logoCopy-link symbol

We are super excited to announce the first-ever Ultralytics YOLOv5 Export Competition with $10,000.00 in cash prizes! Our goal is to help everyone easily train the world's best Vision AI models, and also to help everyone deploy their models just as easily everywhere they want to use them.


The competition will run from May 17th, 2021 to August 31st, 2021.


The deadline for submissions is 24:00 UTC August 31st, 2021. After this date, the competition will be closed and further submissions will not be eligible for prize money.

$10000 in Prizes

The best submission in each of the 5 Categories will claim the full prize funds of $2000.00 (2000.00 USD) from Ultralytics for that category.

5 Categories

Based on community feedback, we created 5 categories which represent the most popular real-world deployment scenarios for YOLOv5 models, including Jetson Nano, Raspberry Pi, Google Edge TPU, desktop CPU and Android edge devices.


To participate, create a public Github repository for your submission, assign your work an open source license and post your submission directly to one of the 5 official EXPORT Competition submissions threads to allow the community to vote. Note these threads are only for official submissions. General questions or comments may be asked directly in this thread, or in a new discussion. Links to submissions:

1. Nvidia Jetson Nano

2. Google Edge TPU

3. Raspberry Pi

4. Intel/AMD CPU

5. Android

Evaluation will take place from September 1, 2021, to September 16, 2021. Winners will be announced in late September 2021, with prizes disbursed immediately thereafter.

Competition Categories

Nvidia Jetson Nano

Evaluation Hardware: Jetson Nano Developer Kit

Prize: $2,000

Google Edge TPU

Evaluation Hardware: Coral Dev Board Mini

Prize: $2,000

Raspberry Pi

Evaluation Hardware: Raspberry Pi 4 Model B

Prize: $2,000


Evaluation Hardware: AWS EC2 t3.medium

Prize: $2,000


Evaluation Hardware: Xiaomi Mi 11

Prize: $2,000

*Funds will be converted to the participant's local currency using the current exchange rate on the date of transfer. Prize money will be transferred via Wise, see the country eligibility list for prize money transfers.


50% of submission scores will be decided by Ultralytics, and 50% will be decided by community feedback, by summing 👍 or 👎 on each submission. Ultralytics scoring will be determined by:

1. Quality of Export (20%)

The simplest export will have the least number of steps, require the least numbers of arguments/parameters, use the least number of imported packages, and be executable with the smallest amount of code.

2. Quality of documentation (20%)

Submissions should be well documented using either a markdown tutorial (i.e. or docs (i.e. Each step should be explained, including setup/requirements, any settings/arguments, export steps, and deployed environment setup, if applicable.

3. Quality of submission (20%)

Every aspect of export and deployment, starting from an official model, should be included. For environments that require special requirements, like Jetson Nano, all packages and/or Docker images must be supplied and documented. For Android deployments, an Android reference app should also be included. A submission must include 100% of what is required to fully export and use a YOLOv5 model.

4. Speed and accuracy of the deployed model (40%)

Deployed models should return near-identical inference results to official YOLOv5 PyTorch models (i.e. inference with python --weights Accuracy of deployed solutions will be analyzed on a hold-out test set of Ultralytics images which are unavailable to the public. Speed is very important also, with the fastest deployment solutions heavily favored. For Android, exports to GPU, NNAPI and Hexagon delegates will receive highest scoring here.

Let’s build the future
of AI together!

Begin your journey with the future of machine learning

Read more in this category