5 ms·

An Introduction to YOLO26

103 points by teleforce 4mo ago

ktallett 4mo ago
I am curious why there is no desire to produce a paper showcasing key details.
- teleforce 4mo ago
  Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models: https://arxiv.org/abs/2606.03748 https://arxiv.org/abs/2606.03748
esquire_900 4mo ago
We've been running YOLO for a number of years (since v5) on soccer videos. None of the recent iterations have been significantly better, with v26 scoring worse then v9 and v11 on our tasks. Makes me wonder why this version is being pushed by roboflow and ultralytics.
- Onavo 4mo ago
  The original YOLO author has long quit due to ethical reasons.
  - utopiah 4mo ago
    Despite having a very memorable paper on the topic I believe they now work at Ai2.
- teruakohatu 4mo ago
  When I was working with YOLO models it did seem like there was little practical improvements were between all of the spinoff models. It seemed people were pushing new models for personal recognition since the original creator stopped working on it. That said, many of the claimed improvements in this model were are efficiency related.
- yfontana 4mo ago
  Can't speak for 26, but a year ago I worked on a project that migrated from v5 to 11 because of improved image segmentation capabilities. My understanding is that the newer versions don't necessarily have better precision/recall, but they tend to be faster for equivalent results, and have increased capabilities.
- GL26 4mo ago
  What I find cool is not the model in itself, but the architectures / training methods found that make the model better. It gives out a new possibilites for other fields of AI. (Notably if you want to fine tune other CV models)
m00dy 4mo ago
Ive used YOLO26 in one of my projects, It was very easy to train on our custom dataset and also very easy to deploy even on rust with AVX2 support. This model is indeed fast and can be used for almost real time inference.
Tepix 4mo ago
With some previous versions of YOLO I‘ve found pages that run it in real-time locally on your browser, analyzing the webcam. Is there a demo like that available for YOLO26?
- GL26 4mo ago
  https://github.com/computer-vision-with-marco/realtime-detection-yolo26 https://github.com/computer-vision-with-marco/realtime-detec...
  - Tepix 4mo ago
    Thanks. Is the result deployed anywhere?
speedgoose 4mo ago
I found that while CLIPSeg is slower than YOLOn, it is still pretty fast and if gave me much much better results without training. If you want to detect objects and speed is important so you can’t use a LLM architecture, you can give it a try too.
yurimo 4mo ago
Wow I'm old, I still remember working with YOLOv2.
geuis 4mo ago
Was evaluating YOLO26 within the last month for its on-device (iPhone 16 Pro) segmentation capabilities. Its decent, but its biggest limitation is that its only trained on 80 COCO classes (meaning pre-labeled images). If whatever is in your images isn't in the 80 classes, its invisible to YOLO26. Conversely I have SAM2 running on-device and its my current workhorse. The biggest benefit with SAM2 for me is that it does fine-grained segmentation masks but isn't trained on labeled images. This was a specific requirement for the app I'm building. SAM2 isn't anywhere as speedy as the native Vision framework apis, but it is more capable across a vastly wider array of potential image targets.
- larodi 4mo ago
  I would prefer GroundingDINo which is a sort of SAM and Dino combo which does open vocabulary.
  - geuis 4mo ago
    Doesn't work for my use-case. GroundingDINO is a text to bounding box model. SAM2 supports coordinate based masks (user taps or clicks somewhere in an image), which is what my research app needs.
Alles 4mo ago
Reminder that Ultralytics is pushing AGPL in a very overreaching way with their models that's why they are not available in Frigate https://github.com/blakeblackshear/frigate/pull/10717 https://github.com/blakeblackshear/frigate/pull/10717
larodi 4mo ago
One thing I don’t get I why the article is credited to ‘Contributing Author’. Meanwhile their very own Peter Skalski already does super job with host write ups and examples of all YOLO sorts and is well respected.
pritambarhate 4mo ago
Is the license for this AGPL? Can someone please confirm?
- trevorhlynn 4mo ago
  Yes AGPL-3.0
pzo 4mo ago
FWIW there are today many more alternatives with better license. Here is a good meta repo for object detection with different model variants: https://github.com/LibreYOLO/libreyolo https://github.com/LibreYOLO/libreyolo
steinvakt2 4mo ago
Just a reminder that RF-DETR is better than yolo26
maelito 4mo ago
Can it measure the speed of a car on a video ?
- MaxikCZ 4mo ago
  Same question, same answer: In pixels/second? Sure! What are you trying to accomplish by those questions? Are you genuinely asking, or just baiting? If the former, didnt answers to your previous question make it clear that your question makes less sense than you might assume?
  - maelito 4mo ago
    It's a serious question. I took a few hours to try this range of models to do this task. Online guides recommend to calibrate the size of the image with a "horizon" line and its real physical size. That's quite complicated. I wish new models coupled with LLM would be capable of estimating the size of features on the map, e.g. the size of the car in meters, to be able to derive the speed with a world understanding. But I have found no resource doing this.
- Joel_Mckay 4mo ago
  Global-shutter cameras are fast and expensive, while Doppler radar modules are robust and under $30 these days. Running machine-vision outside in the Sun or Weather can get tricky. There is also a limited supply of BS a firm can shovel before some bystander ends up dead. =3 https://www.bbc.co.uk/news/articles/c07yp02mxyjo https://www.bbc.co.uk/news/articles/c07yp02mxyjo
alex_duf 4mo ago
I'm sure the model is capable, but I find it funny that the sample image that contains three bears gets detected as two elephants.
- speedgoose 4mo ago
  It’s an accurate representation of the model capabilities in my experience.
deviation 4mo ago
My buddy has some vision impairments, and I remember training a much older of YOLO's models to detect objects/enemies in Terraria for him. It worked very well. I then tried trained it on a lot of sample images from a 3D point & shoot game, and was quite disappointed in how it performed. Has anyone else experimented with it recently? How does this suit as a base-model for training custom classifiers? And with hardware growth in the last ~5 years, is it suitable to run in parallel with games which are graphically intensive?
yeldarb 4mo ago
It’s a big improvement if you’re already paying them but, given their aggressive approach to licensing, I can’t imagine why anyone would choose to use an Ultralytics model on a new project in 2026. You’re just asking to be shaken down and have to pay off a large bill down the line. RF-DETR is both faster and more accurate and truly open source with an Apache 2.0 license: https://github.com/roboflow/rf-detr https://github.com/roboflow/rf-detr Full disclosure: I’m one of the co-founders of Roboflow (we made RF-DETR, wrote this blog post, and are a sub-licensor of Ultralytics’ models.)
- MrGLaDOS 4mo ago
  “RF-DETR is both faster and more accurate and truly open source with an Apache 2.0 license” Misleading marketing statement. The catch is that for image resolutions >=700x700pixels (most production usecases), the roboflow license is actually PML1.0 instead of Apache2.0 https://github.com/roboflow/rf-detr#license https://github.com/roboflow/rf-detr#license
  - krapht 4mo ago
    > The catch is that for image resolutions >=700x700pixels (most production usecases) Citation needed? 2XL looks like you go up to 800x800 pixel inputs. This isn't the dealbreaker you say it is - all pipelines benefit from thoughtful crop and rescaling before going to inference.
    - MrGLaDOS 4mo ago
      See the url in my comment (search for the term rfdetr-2xlarge). 2XL does indeed go up to 800x800 and has PML1.0 license instead of apache 2.0. Rescaling is fine for some purposes but but not for all. For many domain-specific (often less common and odd dimensioned) objects, downscaling will severely reduce recall. There is a reason that Roboflow slaps a license that is not open source on those specific architectures. In some cases tiled inferencing (for example with https://github.com/obss/sahi https://github.com/obss/sahi ) might do the job.

103 points by teleforce 4mo ago

ktallett 4mo ago

I am curious why there is no desire to produce a paper showcasing key details.

teleforce 4mo ago
Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models: https://arxiv.org/abs/2606.03748 https://arxiv.org/abs/2606.03748

esquire_900 4mo ago

We've been running YOLO for a number of years (since v5) on soccer videos. None of the recent iterations have been significantly better, with v26 scoring worse then v9 and v11 on our tasks. Makes me wonder why this version is being pushed by roboflow and ultralytics.

Onavo 4mo ago
The original YOLO author has long quit due to ethical reasons.
- utopiah 4mo ago
  Despite having a very memorable paper on the topic I believe they now work at Ai2.
teruakohatu 4mo ago
When I was working with YOLO models it did seem like there was little practical improvements were between all of the spinoff models. It seemed people were pushing new models for personal recognition since the original creator stopped working on it. That said, many of the claimed improvements in this model were are efficiency related.
yfontana 4mo ago
Can't speak for 26, but a year ago I worked on a project that migrated from v5 to 11 because of improved image segmentation capabilities. My understanding is that the newer versions don't necessarily have better precision/recall, but they tend to be faster for equivalent results, and have increased capabilities.
GL26 4mo ago
What I find cool is not the model in itself, but the architectures / training methods found that make the model better. It gives out a new possibilites for other fields of AI. (Notably if you want to fine tune other CV models)

m00dy 4mo ago

Ive used YOLO26 in one of my projects, It was very easy to train on our custom dataset and also very easy to deploy even on rust with AVX2 support. This model is indeed fast and can be used for almost real time inference.

Tepix 4mo ago

With some previous versions of YOLO I‘ve found pages that run it in real-time locally on your browser, analyzing the webcam. Is there a demo like that available for YOLO26?

GL26 4mo ago
https://github.com/computer-vision-with-marco/realtime-detection-yolo26 https://github.com/computer-vision-with-marco/realtime-detec...
- Tepix 4mo ago
  Thanks. Is the result deployed anywhere?

speedgoose 4mo ago

I found that while CLIPSeg is slower than YOLOn, it is still pretty fast and if gave me much much better results without training. If you want to detect objects and speed is important so you can’t use a LLM architecture, you can give it a try too.

yurimo 4mo ago

Wow I'm old, I still remember working with YOLOv2.

geuis 4mo ago

Was evaluating YOLO26 within the last month for its on-device (iPhone 16 Pro) segmentation capabilities. Its decent, but its biggest limitation is that its only trained on 80 COCO classes (meaning pre-labeled images). If whatever is in your images isn't in the 80 classes, its invisible to YOLO26. Conversely I have SAM2 running on-device and its my current workhorse. The biggest benefit with SAM2 for me is that it does fine-grained segmentation masks but isn't trained on labeled images. This was a specific requirement for the app I'm building. SAM2 isn't anywhere as speedy as the native Vision framework apis, but it is more capable across a vastly wider array of potential image targets.

larodi 4mo ago
I would prefer GroundingDINo which is a sort of SAM and Dino combo which does open vocabulary.
- geuis 4mo ago
  Doesn't work for my use-case. GroundingDINO is a text to bounding box model. SAM2 supports coordinate based masks (user taps or clicks somewhere in an image), which is what my research app needs.

Alles 4mo ago

Reminder that Ultralytics is pushing AGPL in a very overreaching way with their models that's why they are not available in Frigate https://github.com/blakeblackshear/frigate/pull/10717 https://github.com/blakeblackshear/frigate/pull/10717

larodi 4mo ago

One thing I don’t get I why the article is credited to ‘Contributing Author’. Meanwhile their very own Peter Skalski already does super job with host write ups and examples of all YOLO sorts and is well respected.

pritambarhate 4mo ago

Is the license for this AGPL? Can someone please confirm?

trevorhlynn 4mo ago
Yes AGPL-3.0

pzo 4mo ago

FWIW there are today many more alternatives with better license. Here is a good meta repo for object detection with different model variants: https://github.com/LibreYOLO/libreyolo https://github.com/LibreYOLO/libreyolo

steinvakt2 4mo ago

Just a reminder that RF-DETR is better than yolo26

maelito 4mo ago

Can it measure the speed of a car on a video ?

MaxikCZ 4mo ago
Same question, same answer: In pixels/second? Sure! What are you trying to accomplish by those questions? Are you genuinely asking, or just baiting? If the former, didnt answers to your previous question make it clear that your question makes less sense than you might assume?
- maelito 4mo ago
  It's a serious question. I took a few hours to try this range of models to do this task. Online guides recommend to calibrate the size of the image with a "horizon" line and its real physical size. That's quite complicated. I wish new models coupled with LLM would be capable of estimating the size of features on the map, e.g. the size of the car in meters, to be able to derive the speed with a world understanding. But I have found no resource doing this.
Joel_Mckay 4mo ago
Global-shutter cameras are fast and expensive, while Doppler radar modules are robust and under $30 these days. Running machine-vision outside in the Sun or Weather can get tricky. There is also a limited supply of BS a firm can shovel before some bystander ends up dead. =3 https://www.bbc.co.uk/news/articles/c07yp02mxyjo https://www.bbc.co.uk/news/articles/c07yp02mxyjo

alex_duf 4mo ago

I'm sure the model is capable, but I find it funny that the sample image that contains three bears gets detected as two elephants.

speedgoose 4mo ago
It’s an accurate representation of the model capabilities in my experience.

deviation 4mo ago

My buddy has some vision impairments, and I remember training a much older of YOLO's models to detect objects/enemies in Terraria for him. It worked very well. I then tried trained it on a lot of sample images from a 3D point & shoot game, and was quite disappointed in how it performed. Has anyone else experimented with it recently? How does this suit as a base-model for training custom classifiers? And with hardware growth in the last ~5 years, is it suitable to run in parallel with games which are graphically intensive?

yeldarb 4mo ago

It’s a big improvement if you’re already paying them but, given their aggressive approach to licensing, I can’t imagine why anyone would choose to use an Ultralytics model on a new project in 2026. You’re just asking to be shaken down and have to pay off a large bill down the line. RF-DETR is both faster and more accurate and truly open source with an Apache 2.0 license: https://github.com/roboflow/rf-detr https://github.com/roboflow/rf-detr Full disclosure: I’m one of the co-founders of Roboflow (we made RF-DETR, wrote this blog post, and are a sub-licensor of Ultralytics’ models.)

MrGLaDOS 4mo ago
“RF-DETR is both faster and more accurate and truly open source with an Apache 2.0 license” Misleading marketing statement. The catch is that for image resolutions >=700x700pixels (most production usecases), the roboflow license is actually PML1.0 instead of Apache2.0 https://github.com/roboflow/rf-detr#license https://github.com/roboflow/rf-detr#license
- krapht 4mo ago
  > The catch is that for image resolutions >=700x700pixels (most production usecases) Citation needed? 2XL looks like you go up to 800x800 pixel inputs. This isn't the dealbreaker you say it is - all pipelines benefit from thoughtful crop and rescaling before going to inference.
  - MrGLaDOS 4mo ago
    See the url in my comment (search for the term rfdetr-2xlarge). 2XL does indeed go up to 800x800 and has PML1.0 license instead of apache 2.0. Rescaling is fine for some purposes but but not for all. For many domain-specific (often less common and odd dimensioned) objects, downscaling will severely reduce recall. There is a reason that Roboflow slaps a license that is not open source on those specific architectures. In some cases tiled inferencing (for example with https://github.com/obss/sahi https://github.com/obss/sahi ) might do the job.