Stay in the loop with news about Clarifai. Twitter Facebook

Take our demo for a spin.

Send us an image, and we'll understand it using our model trained on 10,000 categories.

Upload an image

Choose image
By uploading an image, you agree to our terms of use.

Or...

Enter an image URL


Auto-Tagging

Recognize tens of thousands of categories, objects, and tags in any image.

train station
  • train
  • railway station
  • subway
  • railroad
rock concert
  • band
  • stage
  • concert
  • music
  • club
  • rock show
toddler flowers
  • garden
  • kid
  • flower
  • woman
  • child
  • tree
camel shadow
  • beach
  • dubai
  • egypt
  • sand
dog snow
  • dog
  • winter
  • snow
alley italy
  • street
  • alley
  • old
  • mexico
  • italian
exotic food market
  • market
  • food
  • turkish
  • istanbul
rusty door handle
  • metal
  • key
  • door
  • old
  • lock
  • wood
rock climber
  • climb
  • rock climber
  • mountain
  • sport

Similar Images

Find similar images in large uncategorized repositories, using a smart combination of semantic and visual similarity. Great for detecting near-duplicates and visual search.

Input Image:

train station

Similar Images:

similar image
similar image
similar image
similar image
similar image
similar image
similar image
similar image
similar image
similar image
similar image
similar image

Speed

Typical recognition speeds under 100ms, using general-purpose GPU techniques.

Accuracy

Our expertise in deep neural nets helped us achieve the world's best published image labelling results.

Scale

We scale to your demand. Our systems are built on proven infrastructure with automatic scaling to handle processing of millions of images.

World Class Performance

Clarifai is pushing the limits of practical artificial intelligence. Our image recognition systems held the top 5 spots for classifying objects in images in the ImageNet 2013 competition. Optimized systems go even further, achieving lower error rates at 10x the speed with 1/5th of the memory footprint.

ImageNet Classification Error Rates
(Lower is Better)
Clarifai: 2013 Advances (10x faster, 5x less memory)
10.7%
Clarifai: ImageNet 2013 Winning Entry
11.2%
2012 ImageNet Winners
15.3%
2012 Traditional Computer Vision Methods
26.2%

The Team

Matthew Zeiler
Matthew Zeiler, PhD

Matthew studied machine learning and image recognition with the pioneers of deep learning and convolutional neural networks. His research produced the top 5 results in the 2013 ImageNet classification competition.

Adam Berenzweig
Adam Berenzweig, PhD

Adam was a software engineer at Google for 10+ years where he built the music recommender for Google Play. Most recently he worked on Goggles and visual search.

Michal Wolski
Michal Wolski

Michal graduated with a BS in Computer Science from Columbia University in May 2014, where he focused on Machine Learning and Computer Vision. As part of a combined plan program he also completed an Applied Mathematics BA from Queens College with honors in mathematics.

Vinay Anantharaman
Vinay Anantharaman

Vinay previously founded a Fashion-Tech startup, BrandBacker and was an engineer on the Adobe Flash Player. His expertise is in web crawling and building scalable back-end systems.

Our Backers Include