# Keras Reviews **Vendor:** Keras **Category:** [Artificial Neural Network Software](https://www.g2.com/categories/artificial-neural-network) **Average Rating:** 4.6/5.0 **Total Reviews:** 65 ## About Keras Keras is a neural networks library, written in Python and capable of running on top of either TensorFlow or Theano. ## Keras Reviews ### 1. A Versatile Tool for Machine Learning, Computer Vision, and Deep Learning **Rating:** 4.0/5.0 stars **Reviewed by:** Verified User in Renewables & Environment | Small-Business (50 or fewer emp.) **Reviewed Date:** April 07, 2026 **What do you like best about Keras?** It’s a versatile tool for machine learning, computer vision, and deep learning. **What do you dislike about Keras?** It has compatibility issues, and on some machines it can be difficult to install and set up. **What problems is Keras solving and how is that benefiting you?** I use it for machine learning and deep learning modeling. ### 2. Best DL Framework **Rating:** 5.0/5.0 stars **Reviewed by:** Aakash Kumar A. | Data Scientist, Enterprise (> 1000 emp.) **Reviewed Date:** September 13, 2023 **What do you like best about Keras?** keras is one of the prominent deep learning framework, it is easy to implement and provides great a amount of important functionalities which helps developer to achieve maximum accuracy **What do you dislike about Keras?** There is nothing to dislike in keras except few things like it still haven't upgraded with the latest functionalities such as nlp and generative AI which are some important tools nowadays **What problems is Keras solving and how is that benefiting you?** Keras helps me to build exceptional deep neural networks and build best models. Keras makes building the neutral network easier. I don't have to manually write anything. ### 3. User-friendly and effective high-level neural-network API **Rating:** 5.0/5.0 stars **Reviewed by:** Subham S. | Data Scientist, Mid-Market (51-1000 emp.) **Reviewed Date:** August 18, 2022 **What do you like best about Keras?** There are a lot of reasons to like Keras: 1. This open-source deep-learning library is designed to provide fast experimentation with deep neural networks. 2. Keras provides the flexibility to run on top of CNTK, TensorFlow, and Theano. 3. It is focused on being modular, user-friendly, readable, and extensible. 4. Keras provides the power to build deep neural networks using fewer lines of code, and this amazes me the most. 5. Since Keras was adopted and integrated into TensorFlow in mid-2017, we can leverage its power by deploying trained models to production thanks to the TensorFlow Serving framework. 6. Keras has excellent access to reusable code and tutorials, which makes it extremely suitable even for beginners. 7. Since Keras runs on top of TensorFlow, it can be equipped with single or multiple GPUs for faster computations. **What do you dislike about Keras?** There are a few reasons for disliking Keras: 1. Keras is not very customizable on its own. While researching different algorithms or working on multi-dimensional matrices, we still need scikit-learn, OpenCV, or Tensorflow for performing such operations. 2. Sometimes the errors are difficult to debug since finding error logs is difficult. For these reasons alone, Keras is still one of the most popular and favorite libraries for statisticians, data scientists, ML engineers, etc. **What problems is Keras solving and how is that benefiting you?** As a data scientist, I have to very frequently build different state-of-the-art deep learning models, and Keras is one such library without which this task seems impossible. One of my recent projects involved Time-Series Forecasting, which involved building different state-of-the-art Deep Learning models like Transformers, LSTMs, etc., which was done using Keras and Tensorflow. If anyone wants to go through a simple implementation using Keras: https://github.com/SubhamIO/KERAS-on-MNIST-dataset ### 4. Keras has always provided with all the tools for machine learning **Rating:** 5.0/5.0 stars **Reviewed by:** Bassel M. | ADAS Machine Learning Engineer, Mid-Market (51-1000 emp.) **Reviewed Date:** January 31, 2021 **What do you like best about Keras?** I like the simplicity of building neural networks **What do you dislike about Keras?** The difficulty of implementing my customized metric **What problems is Keras solving and how is that benefiting you?** I was training a model for classifying images ### 5. Deep Learning made easy and a wonderful library that does offer a lot! **Rating:** 4.0/5.0 stars **Reviewed by:** Sathesh R. | Managed Services Integration Analyst, Enterprise (> 1000 emp.) **Reviewed Date:** September 15, 2022 **What do you like best about Keras?** Best wrapper API available out there for Neural networks. You need not have to be an expert programmer, it does offer what you need to get the job done and it is an open source. Integrates well with tensor flow. Its native to python and I come up with python background, it does make my coding world a lot easier. Implementing a neural network would take hours of coding, but Keras has made it simpler with few lines of code and it is easily understandable. **What do you dislike about Keras?** Understanding of the log traces to fix a problem takes time, as you would have to understand the way it is traced and written, that would take time because of limited documentation. As a python developer I find it to be easier to use, but it doesnt provide other language support, it could be a problem for a long term development. It doesnt offer a great backend support as it is limited. **What problems is Keras solving and how is that benefiting you?** It does make development faster by offering a lot with neural networks, it makes the neural network implementation much easier and smoother which helps save lot of time. ### 6. Open source tool for managing artificial neural networks **Rating:** 4.0/5.0 stars **Reviewed by:** Paweł W. | Software Engineer @ Creators, Small-Business (50 or fewer emp.) **Reviewed Date:** October 09, 2022 **What do you like best about Keras?** First of all Keras is a complete API for managing neural networks and is an open source tool. I find its API extremely convenient to use - definitely simpler to use than PyTorch **What do you dislike about Keras?** It might get slow for some complicated use cases, so if you are aiming for speed and efficiency then probably PyTorch would be a better choice. **What problems is Keras solving and how is that benefiting you?** Keras provides an API that allows me as a developer to avoid some of the typical boilerplate code that would be written without using any framework. It lets me focus on modelling the problem rather than reinventing the wheel ### 7. User friendly high level framework for developing deep neural networks **Rating:** 4.5/5.0 stars **Reviewed by:** Dipak K. | Senior Research Fellow (PhD), Small-Business (50 or fewer emp.) **Reviewed Date:** June 08, 2022 **What do you like best about Keras?** easy and quick implementation of a variety of neural network models. simple and easy to learn with vast support from the Keras community and documentation. I like the most about Keras is the high-level framework and runs on top of TensorFlow with single or multiple GPUs for faster computations. availability of pre-trained models such as VGGNET, RESNET, etc. **What do you dislike about Keras?** Preprocessing of the signals or images is still not widely used due to lack of customization. one needs to use additional tools such as Scikit-learn to do the proper preprocessing. Issues in the low-level backend can not be targeted and finding those error logs is difficult. Other than these issues Keras is widely famous in the AI field. **What problems is Keras solving and how is that benefiting you?** Building state-of-the-art neural network models for various applications such as classification and identification tasks in the biomedical field. Few deep learning model examples are ResNET, Alexnet, autoencoders etc. ### 8. An excellent neural network library to run our models **Rating:** 4.0/5.0 stars **Reviewed by:** Ashish T. | System Engineer, Mid-Market (51-1000 emp.) **Reviewed Date:** September 27, 2022 **What do you like best about Keras?** It is easier to use and setup on most of the backend systems like tensor flow and pytorch. This provides a lot of operational freedom to developers to experiment. **What do you dislike about Keras?** Some external integration is difficult to implement on the system and requires assistance from consultants. Initial setup on windows OS is also a bit challenging. **What problems is Keras solving and how is that benefiting you?** Keras is helping in getting our deep learning projects off the ground, it is very helpful in running these projects and allows us to maintain and run the models . ### 9. Great for beginners and for somewhat advanced as well. **Rating:** 4.0/5.0 stars **Reviewed by:** Gourav S. | Open Source Contributor, Small-Business (50 or fewer emp.) **Reviewed Date:** September 13, 2022 **What do you like best about Keras?** Keras is amazing with its documentation and I've used it on Google collab. It worked super well, models were meeting expectations. **What do you dislike about Keras?** Might not be that great as compared to alternatives, when it comes to speed, it's somewhat slow. **What problems is Keras solving and how is that benefiting you?** It's good if you wanna learn Tensorflow and Tensorflow 2. As it's built on following Api. ### 10. Makes deep learning easy **Rating:** 4.0/5.0 stars **Reviewed by:** Shubham C. | DECISION ANALYTICS ASSOCIATE, Small-Business (50 or fewer emp.) **Reviewed Date:** September 01, 2022 **What do you like best about Keras?** Keras makes deep learning easy . Its easy to use and every code is thoroughly explained in the website **What do you dislike about Keras?** Codes should be more easy to find. apart from that there is no issue **What problems is Keras solving and how is that benefiting you?** I wanted to learn deep learning. Being from a non technical Background keras has made it easy for me ### 11. Keras, Best Deep Learning Framework with optimisers **Rating:** 4.5/5.0 stars **Reviewed by:** poorna c. | Senior Engineer, Mid-Market (51-1000 emp.) **Reviewed Date:** June 08, 2022 **What do you like best about Keras?** The most liked feature of Keras is it wraps the large chunks of codes in inbuilt functions, it is easy to write or implement the ANN compare to TensorFlow, Well documented. **What do you dislike about Keras?** Overall Keras is good and has not many drawbacks as such, the only thing which can be improved in Keras is its performance on large numbers of epochs or iterations while training the model. **What problems is Keras solving and how is that benefiting you?** We are developing and implementing the deep learning neural networks using Keras, we are developing the auto face recognition engine for the security cameras in premises. ### 12. Keras - Open Source Neural Network Platform **Rating:** 4.5/5.0 stars **Reviewed by:** Senthil N. | Software Engineer, Small-Business (50 or fewer emp.) **Reviewed Date:** September 09, 2022 **What do you like best about Keras?** Keras is the best platform that runs in many venues. Like TensorFlow, Microsoft Cognitive Services etc... **What do you dislike about Keras?** Not suitable for beginners who need initial setup and more technical knowledge in Tensorflow. **What problems is Keras solving and how is that benefiting you?** Keras helps to train and test computer vision and NLP models ### 13. Best ML Framework **Rating:** 5.0/5.0 stars **Reviewed by:** Vineet T. | Student at IIT Indore, Small-Business (50 or fewer emp.) **Reviewed Date:** September 13, 2022 **What do you like best about Keras?** Easy to use, very flexible, easy for deployment **What do you dislike about Keras?** Poor support for Quantum machine learning **What problems is Keras solving and how is that benefiting you?** Functional API for research problems ### 14. Excellent library and interface for TensorFlow **Rating:** 5.0/5.0 stars **Reviewed by:** Verified User in Research | Mid-Market (51-1000 emp.) **Reviewed Date:** September 13, 2022 **What do you like best about Keras?** - Easy to use - Can code up deep learning models quickly **What do you dislike about Keras?** - Advanced customization may not be easily achieved **What problems is Keras solving and how is that benefiting you?** - Deep learning library for various tasks like image recognition, NLP, and more ### 15. The best way of accessing tensorflow **Rating:** 4.5/5.0 stars **Reviewed by:** Oliver G. | Technical Sales Engineer, Enterprise (> 1000 emp.) **Reviewed Date:** June 20, 2022 **What do you like best about Keras?** Getting results with neural networks very wuickly **What do you dislike about Keras?** Incompatibility when using Keras and Keras within tensorflow **What problems is Keras solving and how is that benefiting you?** Get very good results by letting me flexibly build networks ### 16. Framework You Need for AI **Rating:** 4.5/5.0 stars **Reviewed by:** Chandresh M. | System Engineer, Mid-Market (51-1000 emp.) **Reviewed Date:** October 08, 2021 **What do you like best about Keras?** The first thing I like about Keras is that it is a very user-friendly API. I can create Deep Learning models very quickly because of their very rich functions. Another important thing is its community. Whenever I find any problem in code then I can look for it in the Keras community; because of its popularity, you can find solutions a little bit easier. It also provides many pre-trained models like Xception, MobileNet, VGG16, InceptionV3, and many more. So I can use these models without training. If you have GPUs, then it also provides GPU support, which makes training of models faster. **What do you dislike about Keras?** What I don't like about the Keras is that it doesn't have good functions for data processing. It also needs some improvements. Sometimes it takes more time to trained models because of backend issues. **Recommendations to others considering Keras:** If you have some knowledge about Machine Learning and Deep Learning, then you can use Keras. **What problems is Keras solving and how is that benefiting you?** I am using Keras for developing Machine Learning and Deep Learning classification models. ### 17. It was really helpful in most of my projects and thesis work. **Rating:** 3.5/5.0 stars **Reviewed by:** Ashin Marin T. | Software Development Engineer, Enterprise (> 1000 emp.) **Reviewed Date:** January 16, 2021 **What do you like best about Keras?** The support and Keras dependability with Anaconda. **What do you dislike about Keras?** Sometimes while training on keras, it gets stuck. The other issue is when installing other softwares on keras, it crashes. **What problems is Keras solving and how is that benefiting you?** Problem realted to image classification and facial expression detection. ### 18. Easy and successful **Rating:** 5.0/5.0 stars **Reviewed by:** deniz y. | Business Intelligence Manager, Enterprise (> 1000 emp.) **Reviewed Date:** October 02, 2021 **What do you like best about Keras?** Neural networks are indispensable for data science and machine learning. It is quite easy to use if you have a little knowledge of the math of the job. The documentation and other resources are good and satisfying. The examples on keras.io are very instructive. The image processing module is successful. Runs models smoothly on GPU and CPU. It supports RNN and CNN. **What do you dislike about Keras?** Overall, I'm satisfied, but I feel like I'm not in full control of the model. **What problems is Keras solving and how is that benefiting you?** I use it for text and image classification. I analyze the data quickly and automatically by classifying it with high accuary. ### 19. Keras deep learning API **Rating:** 4.5/5.0 stars **Reviewed by:** GOURI S. | Technical Lead Data Scientist, Mid-Market (51-1000 emp.) **Reviewed Date:** November 09, 2021 **What do you like best about Keras?** What I liked the most about Keras framework is unlike tensorflow it provide easy set of code lines, using that we can develop deep learning model easily. **What do you dislike about Keras?** What I liked the least about Keras framework is it provides errors sometimes those are tough to debug. **What problems is Keras solving and how is that benefiting you?** I have used Keras library to train my deep learning model ( Artificial neural networks). It provides better accurate trained models. ### 20. Best framework for people who start with deep learning **Rating:** 5.0/5.0 stars **Reviewed by:** Sai Vignan M. | P, Mid-Market (51-1000 emp.) **Reviewed Date:** July 25, 2021 **What do you like best about Keras?** 1. Very easy to implement 2. very easy to learn 3. has many parameters/options to trigger algorithms 4. Good for deep learning starters **What do you dislike about Keras?** 1. Not so advanced and debuggable compared to tensorflow or pytorch 2. Not so fast in computation compared to other frameworks **Recommendations to others considering Keras:** Best library for starters **What problems is Keras solving and how is that benefiting you?** solved basic Deep learning usecases using methods of LSTM, CNN. A lot of usecases are used for this. 90% of deep learning enthusiasts use this. We use for NLP and ANN in general ### 21. Keras is a practical, easy to use package **Rating:** 5.0/5.0 stars **Reviewed by:** Phuong N. | Data Scientist, Small-Business (50 or fewer emp.) **Reviewed Date:** January 27, 2021 **What do you like best about Keras?** Keras is very easy to use, even for beginners that have basic Python programming skills. Even complex deep learning models can be built just with a few lines of codes. The biggest advantage is running time: the codes execute pretty fast. Besides, code examples are intuitives and readily availables. The documentation is built with care and attention and there are answers for almost every issues. I always find what I need to solve my problem. **What do you dislike about Keras?** Sometimes it is not easy to find code examples for some advanced features. Also there were some errors in the code execution and I had difficulty to understand where they came from. Since Keras is pretty simple code, sometimes it is hard to customize models that have been built by someone else. In such case, I would rather use other packages that might be more complex but do the job. **Recommendations to others considering Keras:** I would say: let's give it a try. I first knew about Keras when I was a Python beginner and was immediately impressed by its user-friendly aspect. Besides that, the documentation is complete and the community is there to help. **What problems is Keras solving and how is that benefiting you?** I used Keras to build several deep learning models in different topics: image classification, time series prediction, categorical variable prediction, object detection in images. I used it in both my academic research as well as in professional projects. I successfully built a CNN combined with a RNN model that was later deployed in a mobile application. Even though the original dataset was big, Keras has an impressive running time. That helped me a lot to speed up the execution time of my codes, which is crucial for the project's success. ### 22. Great way to take new steps in Deep Learning! **Rating:** 4.0/5.0 stars **Reviewed by:** Argyrios L. | Sales Administrator, Mid-Market (51-1000 emp.) **Reviewed Date:** January 22, 2021 **What do you like best about Keras?** Easy of use when it comes to model creation and implementation. In general I like of how naturally the Keras methods are executed like original python methods. Overall, the focus on user experience is what makes Keras approachable by interns like me. **What do you dislike about Keras?** I did not like how I needed to install so many pre-requisites to work on a relatively simple project that I had in mind, however that's understandable provided how complex library dependencies can get in software engineering. **Recommendations to others considering Keras:** If you're planning to use Keras for educational and small scale projects, it's the way to go! **What problems is Keras solving and how is that benefiting you?** The project I worked on was face mask recognition. The project revolves around recognizing whether a person wears a face mask or not. It seemed like an interesting project at the beginning, however it took more time to setup and get it running than I anticipated. The benefits I got thus far was a brief introduction to what deep learning can offer if one invests enough time on it. On the future I would like to improve on the project by adding an extra feature so that the model can detect whether a person doesn't have their nose covered or not, however I will need more material/data for this (more pictures of people wearing their mask incorrectly) ### 23. It has allowed me to create the necessary code for my neural network applications very easily **Rating:** 4.0/5.0 stars **Reviewed by:** Honker B. | data engineer, Mid-Market (51-1000 emp.) **Reviewed Date:** January 14, 2021 **What do you like best about Keras?** Keras turns out to be a fairly complete project to achieve neural networks in a fairly simple way, the ease with which it can be executed and its code turns out to be quite impressive, especially because despite being a specialized use program , a person with basic knowledge in the area of ​​programming may be perfectly capable of making a neural network without so many complications, the simplicity of its framework is something that is quite impressive, especially because it allows me to do everything in a much more practical way and efficient and without the existence of so many complications. **What do you dislike about Keras?** If there is an error in the configuration or syntax of the code, it is quite problematic to locate it to be able to change it and fix the operation of the entire system, it is something quite cumbersome to do so they should implement error detection mechanisms as a suggestion . **What problems is Keras solving and how is that benefiting you?** It has allowed me to create the necessary code for my neural network applications in a quite simple and simple way so that my projects can be executed without so many complications or variations, its simplicity is something quite impressive, not to mention the fact that it has infinities of applications that are not only limited to the development of neural networks, but since it is based on Python as a programming language, it also allows us the possibility of using it in applications other than networks. ### 24. Easy to learn and customisable **Rating:** 5.0/5.0 stars **Reviewed by:** Vaishak K. | Deep Learning Consultant, Small-Business (50 or fewer emp.) **Reviewed Date:** January 22, 2021 **What do you like best about Keras?** Keras has a very simple api which is easy to learn for Machine Learning practitioners. Whether using an in-built model or building a custom model, Keras is very intuitive to use. It is a must have package for any Data Scientist/ ML Engineer. **What do you dislike about Keras?** Keras miss some good features of Fastai like learning rate finder and one-cycle learning out of the box. **What problems is Keras solving and how is that benefiting you?** I'm mostly building deep learning solutions for Computer Vision problems using Keras. With Keras it is very convenient to customise and experiment with different model configurations. ### 25. Very good experience working with Keras **Rating:** 5.0/5.0 stars **Reviewed by:** Sadi A. | Software Engineer(IOS), Small-Business (50 or fewer emp.) **Reviewed Date:** February 05, 2021 **What do you like best about Keras?** The API in Keras is super user friendly. Everything is integrated right into the platform with a Tensorflow backend. **What do you dislike about Keras?** Since the platform was new when i used it, there were not many customization options available to the deep learning model that i was building. **What problems is Keras solving and how is that benefiting you?** I was training a Convolutional Neural Network to understand and learn from images. The implementation was very easy in Keras platform compared to Pytorch or Tensorflow. ### 26. Great interface and good performance **Rating:** 4.5/5.0 stars **Reviewed by:** Elena R. | Computer Vision Engineer , Mid-Market (51-1000 emp.) **Reviewed Date:** January 28, 2021 **What do you like best about Keras?** It is very easy to use, while guaranteeing great performance. Perfect for beginners and to implement different state of the art networks without big effort. **What do you dislike about Keras?** The fact that is a high-level interface makes some features to work in a black box manner. **What problems is Keras solving and how is that benefiting you?** I used it to solve problems related to deep learning and reinforcement learning. ### 27. My Experience With Keras **Rating:** 5.0/5.0 stars **Reviewed by:** Verified User in Higher Education | Small-Business (50 or fewer emp.) **Reviewed Date:** January 28, 2021 **What do you like best about Keras?** I used Keras at university in my AI module and I found it very useful as it is a great tool to use alongside NLP. **What do you dislike about Keras?** At first it was difficult to use as I was a beginner. So I guess that’s one thing, it could’ve had a simpler methodology to it. **What problems is Keras solving and how is that benefiting you?** I was able to create a Chatbot where Keras played a big part of solving the problem. Without Keras it wouldn’t have been possible to build the Chatbot. ### 28. Keras makes it easier and quicker to write deep learning models **Rating:** 5.0/5.0 stars **Reviewed by:** Sudarshan P. | Postdoctoral Researcher, Mid-Market (51-1000 emp.) **Reviewed Date:** January 16, 2021 **What do you like best about Keras?** I like the way keras helps to write complex models easily with few lines of code, and the implementations of different network architecture make it so convenient to work. **What do you dislike about Keras?** Sometimes it is difficult to customize as required. **What problems is Keras solving and how is that benefiting you?** I am working in different problems mostly related images. ### 29. I am using Keras for my deep learning projects. **Rating:** 4.5/5.0 stars **Reviewed by:** Verified User in E-Learning | Small-Business (50 or fewer emp.) **Reviewed Date:** August 11, 2020 **What do you like best about Keras?** It is user friendly and easy to learn. It runs on TensorFlow and TensorFlow is the most popular software library for deep learning these days. **What do you dislike about Keras?** Since it is a high-level library, sometimes slower than Tensorflow. **Recommendations to others considering Keras:** If you are new to deep learning and want to learn to code, Keras is a good start, because it is user-friendly and easy to learn API. **What problems is Keras solving and how is that benefiting you?** I am learning Deep learning and using Keras for deep learning models. Because of its user-friendliness, we can build a model and train it with a few lines of codes. Also, Keras has GPU support and it is very useful for speeding up the training process of network models. ### 30. Keras facilitates the usage of different machine learning/deep learning backend libraries **Rating:** 5.0/5.0 stars **Reviewed by:** Verified User in Information Services | Small-Business (50 or fewer emp.) **Reviewed Date:** January 30, 2021 **What do you like best about Keras?** The simple to understand and easy to use syntax **What do you dislike about Keras?** At this moment, i do not have anything in particular **What problems is Keras solving and how is that benefiting you?** I am using Keras for creating deep learning architectures ### 31. Keras is great user friendly product **Rating:** 4.5/5.0 stars **Reviewed by:** Jay P. | Data Scientist - Big Data and Analytics, Automotive, Enterprise (> 1000 emp.) **Reviewed Date:** September 01, 2019 **What do you like best about Keras?** The best part is if offers broad adoption, support for a wide range of production deployment options, integration with at so our best framed back-end engines - super easy to deploy **What do you dislike about Keras?** We think Keras should do its own operations - not rely on back-end services. I know its too much to ask for but it would be great to have it's own. Just that too much reliability on low level operations makes us think - did they do that due to infrastructure or was it a choice? That being said - the feature extraction is something to die for - with our favorite language Python in use **Recommendations to others considering Keras:** Well I recieved no Keras training - I had to figure out all on my own but documentation was good - I would advice that the community for Keras to have more Youtube videos to explain the deployment to cloud. I could not find anything relevant to what I was looking for. Maybe the deployment to our server doesn't allow cloud integration - I am not aware **What problems is Keras solving and how is that benefiting you?** We were using Google cloud for our company (FCA Automobiles) to understand image classification (we used VGG19) to separate out Vehicles with poor torque alignments in Manufacturing line. The images are provided by inbuilt robotic arms and we use Keras as software to understand which vehicles are pushed in for quality checks. We have had immense quality improvements done and line speed changed when we realized how change of torque could help us in better images ### 32. Tensorflow makes huge improvement by having a canonical format supported by all projects **Rating:** 5.0/5.0 stars **Reviewed by:** Chris H. | Data Scientist/Video Analytics/Software, Computer Software, Enterprise (> 1000 emp.) **Reviewed Date:** October 21, 2019 **What do you like best about Keras?** I like that tensorflow is moving away from tensorflow extended, slim, keras, etc. formats to having everything be done in keras. It is see hardware that touts support for tensorflow models only to learn that the associated software only supports object detection api models or only supports tensorflow extended models. I want things to just work across hardware and software libraries. **What do you dislike about Keras?** Although Keras is an improvement towards easy customization of existing models it doesn't go far enough in making it easy to try out a variety of custom architectures. Simple things like being able to view weights on a neuron and surgically remove it if it doesn't meet some custom criteria, then save and reload are not supported. I'm using third party projects to try to make this easier (https://github.com/Raukk/tf-keras-surgeon/), but I feel like the third party project is a bit of a hack and could be improved by having native library support. It is hard to take a new an improved model that someone else has written and tweak it with the method from a recent paper of replacing all the X layers with Y, for which they saw some improvement. I have about twenty papers with improvements over a baseline model and a lot of these aren't independent, but putting them all into one model is unnecessarily difficult. **What problems is Keras solving and how is that benefiting you?** We have been working to create new object detection models using state-of-the-art classification models. Specifically efficientnets (https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet) by default only support classification. We are interested in object detection. The tensorflow 1.0 has an object detection api, which is great for quickly creating streamlined object detection models, but it doesn't play nice with tensorflow extended models, keras models, or other tensorflow models. We are working on creating a tensorflow 2.0 keras object detection model, so that we can easily swap out the backbone or feature extractor from the object detection model to get a new more accurate object detection model. ### 33. Keras for fast and easy image classification **Rating:** 5.0/5.0 stars **Reviewed by:** Stanley D. | Data Engineer, Computer Hardware, Small-Business (50 or fewer emp.) **Reviewed Date:** June 27, 2019 **What do you like best about Keras?** 1. I love the fact that with very few lines of code and little knowledge about deep learning and convolutional neural networks, one can easily build an image classifier using by just reading its documentation. 2. Keras community and developer support is very high and active, so there are a lot of tutorials on Keras available. 3. I love the fact that I can easily integrate Keras and Scikit-learn, and apply functions on scikit-learn such as ensembles, cross-validation and stacking. **What do you dislike about Keras?** Keras is a high-level deep learning framework, which makes customization and tweaking hard. **What problems is Keras solving and how is that benefiting you?** When I was new to deep learning, I was able to build an image classifier using Keras in a deep learning competition by just reading its documentation ### 34. A beautiful Deep Learning Framework **Rating:** 5.0/5.0 stars **Reviewed by:** Verified User in Computer Software | Mid-Market (51-1000 emp.) **Reviewed Date:** July 15, 2019 **What do you like best about Keras?** Keras is an open source deep learning framework with lots and lots of features it allows you to do so many things like creating multi later neural networks etc. Keras is performs computations quickly and it is built upon Tensorflow which is one of the best frameworks out there. The fact I can build a full neural network in less than 9 lines is because of keras. It also has all the required deep learning functions killer Dropout,Dense, softmax amongst others **What do you dislike about Keras?** Keras is structured as an API to simply make function calls to Tensorflow. Making it hard to implement features that aren't out of the box **Recommendations to others considering Keras:** If you need a reliable deep learning framework to give you results that's also easy to work with, keras is the way to go **What problems is Keras solving and how is that benefiting you?** Keras has allowed me to rapidly develop and test out my deep learning models. This means that I have more work output and can be more productive ### 35. I love Keras **Rating:** 5.0/5.0 stars **Reviewed by:** Anusha K. | Data Scientist, Enterprise (> 1000 emp.) **Reviewed Date:** October 18, 2019 **What do you like best about Keras?** Keras is very flexible abd user-friendly when it comes to training models **What do you dislike about Keras?** Not compatible with Tensorflow. Unable to convert some trained TF models to Keras **Recommendations to others considering Keras:** Easy to use, flexible. Start from Keras. Switching from other softwares after training is a bad idea, might have to implement models from scratch again, not worth the time and effort **What problems is Keras solving and how is that benefiting you?** Computer Vision model training and evaluation. ### 36. Keras is one of the best **Rating:** 5.0/5.0 stars **Reviewed by:** ahmed i. | Computer Vision Researcher, Enterprise (> 1000 emp.) **Reviewed Date:** August 20, 2019 **What do you like best about Keras?** I believe Keras is one of the main pillars that raised deep learning to where it is now. It is easy to use and powerful. The seamless integration with Tensorflow is fantastic. Keras is one of the best if not the best. **What do you dislike about Keras?** The only thing I did not like during my journey with Keras is how they did not do a good job of keeping backward compatibility. Sometimes, I had to make major changes to accommodate newer versions. **What problems is Keras solving and how is that benefiting you?** I used Keras to do most of my PhD research work. It did help me do my research faster enabling me to perform more experiments in the same amount of time. ### 37. Oh hello tensorflow 2.0 finally you see keras is important **Rating:** 4.5/5.0 stars **Reviewed by:** Pavan C. | Software Engineer Intern, Enterprise (> 1000 emp.) **Reviewed Date:** July 05, 2019 **What do you like best about Keras?** Its the frontend every ML library should use. The api is quite understandable, building up layers makes sense, the estimator api of tensorflow is too lost in the details to abstract such ease of use. Loss functions, activations, optimizers, lstm you got it all **What do you dislike about Keras?** Number of open issues on github, this project needs more directed support. It can rock the AI world. Tensorflow realized its importance and optimized it to intergrate it into the roots in 2.0 but still 2.0 is not released on stable channel. A few months more maybe **Recommendations to others considering Keras:** Get done with initial learning curve, you'll love it later **What problems is Keras solving and how is that benefiting you?** neural networks that are super optimized and built with ease ### 38. Keras: An excellent abstraction of Tensorflow **Rating:** 4.5/5.0 stars **Reviewed by:** Abel A. | Software Engineer, Enterprise (> 1000 emp.) **Reviewed Date:** September 02, 2019 **What do you like best about Keras?** I love the ease that Keras provides, offering a brilliant abstracted wrapper around TensorFlow, allowing you to build most customizations for neural networks. **What do you dislike about Keras?** The fact that Keras is dropping functionalities for other frameworks one after the other and proving to be a strong abstraction of TensorFlow. What is the difference between TensorFlow Keras then? **Recommendations to others considering Keras:** Use TF.keras and then explore how the underlying G framework was constructed, and compare it to Torch. **What problems is Keras solving and how is that benefiting you?** Simple, fast prototyping of machine learning networks for customized solutions. ### 39. Great for intermediate users **Rating:** 4.0/5.0 stars **Reviewed by:** Benoît B. | Data Scientist / Machine Learning Engineer, Small-Business (50 or fewer emp.) **Reviewed Date:** September 03, 2019 **What do you like best about Keras?** easy to deploy models, fine tune them, and export them **What do you dislike about Keras?** the documentation is lacking; only the most basic of networks are clearly explained; the concepts behind the hyperparameters are not explained at all (what is an optimizer? what is the difference between adagrad and adam?); how do you build more complicated networks like recurrent NN ? **Recommendations to others considering Keras:** Make sure to read the documentation, starting with the examples. **What problems is Keras solving and how is that benefiting you?** We are solving Human Activity Recognition. ### 40. Have a great experience **Rating:** 5.0/5.0 stars **Reviewed by:** Verified User in Information Technology and Services | Enterprise (> 1000 emp.) **Reviewed Date:** October 02, 2019 **What do you like best about Keras?** Keras is a high level API. It is easy to use. Better than tensorflow (low level api). **What do you dislike about Keras?** There is nothing in keras which you don't like. It is perfect. **Recommendations to others considering Keras:** Keras is easy to use , if you want to make artificial intelligence model. **What problems is Keras solving and how is that benefiting you?** I work in keras and tensorflow. I solve Artificial intelligence problems with them. ### 41. Keras - Convolutional Neural Netwroks **Rating:** 4.0/5.0 stars **Reviewed by:** Verified User in Information Technology and Services | Enterprise (> 1000 emp.) **Reviewed Date:** August 30, 2019 **What do you like best about Keras?** If you want to quickly build and test a neural network with minimal lines of code, Keras is the best framework. With Keras, you can build simple or very complex neural networks within a few minutes. Keras is more user friendly. **What do you dislike about Keras?** If someone wants more control over your network and want to watch closely what happens with the network over the time, Tensor Flow is the right choice. Keras has been integrated in TF, it is wiser to build your network using tf.keras and insert anything you want in the network using pure TensorFlow. **What problems is Keras solving and how is that benefiting you?** Implementing a Convolutional Neural Network Implementing a Recurrent Neural Network Functional API ### 42. easy to understand and use **Rating:** 4.5/5.0 stars **Reviewed by:** Verified User in Information Technology and Services | Small-Business (50 or fewer emp.) **Reviewed Date:** September 09, 2019 **What do you like best about Keras?** To understand deep learning training and layer management, keras helped me a lot. Very easy to manage, change, manipulate the layers. Parameters are also easy to tune. **What do you dislike about Keras?** There are some fairly advanced frameworks out there compared to keras. It definitely helps a beginner. But I wish it could be used in a more advanced level as well. **Recommendations to others considering Keras:** At the beginner level, Keras is really good to understand how deep learning works. **What problems is Keras solving and how is that benefiting you?** I used keras for a text classification task. I wanted to see how the model performs on different types of layers. And it was quite easy to modify. ### 43. Keras is simple and straightforward **Rating:** 4.0/5.0 stars **Reviewed by:** Verified User in Research | Mid-Market (51-1000 emp.) **Reviewed Date:** September 26, 2019 **What do you like best about Keras?** It is very easy to prototype new research ideas using Keras. The support for multi GPU and distributed training is very good. **What do you dislike about Keras?** Not native to python, so makes it slightly hard to debug sometimes. Different backends are not robust often. **What problems is Keras solving and how is that benefiting you?** I used Keras to build deep learning models to do image classification task. To be specific I used for whether classification for ADAS project. ### 44. Useful, high level abstraction **Rating:** 3.5/5.0 stars **Reviewed by:** Rejin J. | Senior Staff Computer Vision Engineer, Computer Software, Mid-Market (51-1000 emp.) **Reviewed Date:** August 26, 2019 **What do you like best about Keras?** Easy to whip up a small network without worrying about dimensionality matching **What do you dislike about Keras?** Not enough control over graphs/loss functions/implementations **What problems is Keras solving and how is that benefiting you?** Object tracking prototyping ### 45. Simple and easy to use **Rating:** 4.0/5.0 stars **Reviewed by:** Verified User in Health, Wellness and Fitness | Small-Business (50 or fewer emp.) **Reviewed Date:** August 16, 2019 **What do you like best about Keras?** Sprint model creation, easy to create and test simple models, automatic early stoppi g and easy to use data generators. **What do you dislike about Keras?** Model transfer use from multi gpu to single gpu **Recommendations to others considering Keras:** If I want to use advance features like TFRecords, I will be better using tensorflow, so automating from input to output for traning using TFrecord would be a good addition **What problems is Keras solving and how is that benefiting you?** Digital pathology ### 46. The Best Platform for Deep Learning and NN **Rating:** 4.0/5.0 stars **Reviewed by:** Verified User in Telecommunications | Enterprise (> 1000 emp.) **Reviewed Date:** August 30, 2019 **What do you like best about Keras?** The best thing that I like about Keras is that it is easy and convenient to use. It is a Unified platform **What do you dislike about Keras?** I am still exploring it more, will know after a few **What problems is Keras solving and how is that benefiting you?** I am using Keras for Deep Learning models like RNN, Dense, LSTM Neural Network on Sequential Time Series Data ### 47. Very Useful Tensorflow Wrapper **Rating:** 5.0/5.0 stars **Reviewed by:** Verified User in Information Technology and Services | Small-Business (50 or fewer emp.) **Reviewed Date:** February 06, 2019 **What do you like best about Keras?** Keras drastically simplifies building neural networks in Tensorflow without losing too much customizability/functionality. I prefer Keras over TF always unless I know for sure that the best way to use some aspect is within TF itself. **What do you dislike about Keras?** For beginners, bypassing concepts like matrix multiplication/dimensions can make some errors hard to rectify. So while it's a great way to get your feet wet building NNs, some coding with Tensorflow can be useful to keep your shapes in line. **Recommendations to others considering Keras:** If you are not already familiar with matrix multiplication and dimensions/shapes, I recommend familiarizing yourself with these concepts in TF before jumping into Keras. **What problems is Keras solving and how is that benefiting you?** I primarily use Keras to build neural nets for financial prediction and natural language processing. ### 48. My honest review about Keras. **Rating:** 4.0/5.0 stars **Reviewed by:** Verified User in Education Management | Mid-Market (51-1000 emp.) **Reviewed Date:** August 19, 2019 **What do you like best about Keras?** Model building using Keras is super easy for beginners. The fact that it uses TensorFlow as it’s backend for model building and execution is interesting because it makes the model efficient. **What do you dislike about Keras?** There is nothing that I currently dislike about Keras. **What problems is Keras solving and how is that benefiting you?** I mostly use Keras for building models to solve problem in the domain of Computer Vision. ### 49. Keras - the easiest to learn AI/ML framework **Rating:** 4.5/5.0 stars **Reviewed by:** Verified User in Information Technology and Services | Enterprise (> 1000 emp.) **Reviewed Date:** August 26, 2019 **What do you like best about Keras?** Easy to learn with many options and lots of examples. You can quickly find test many models or even different approaches very quickly **What do you dislike about Keras?** Hard to find documentation on some of the more obscure options that are dependent on other frameworks. **Recommendations to others considering Keras:** Try and learn **What problems is Keras solving and how is that benefiting you?** Image classification, NLP, text analytics ### 50. Keras is one of the fastest DL frameworks ever **Rating:** 5.0/5.0 stars **Reviewed by:** Verified User in Internet | Mid-Market (51-1000 emp.) **Reviewed Date:** September 03, 2019 **What do you like best about Keras?** I really enjoy working in Keras. Its fast and efficient and is easily doable. **What do you dislike about Keras?** Keras layers can only return a single tensor. For building a moderately complex network, this can result is lot of extra code which can be time consuming **What problems is Keras solving and how is that benefiting you?** Fast and powerful ## Keras Discussions - [what's the best way to go about the documentation of Keras . lack of clarity is provided on various topics for example cooldown parameter in ReduceLROnPlateau is clearly not explained .](https://www.g2.com/discussions/36779-what-s-the-best-way-to-go-about-the-documentation-of-keras-lack-of-clarity-is-provided-on-various-topics-for-example-cooldown-parameter-in-reducelronplateau-is-clearly-not-explained) - 1 upvote - [How can I use Keras for text recognition?](https://www.g2.com/discussions/36633-how-can-i-use-keras-for-text-recognition) - 1 upvote - [Is there any work going on to make conversions from Tensorflow to Keras and vice versa possible?](https://www.g2.com/discussions/is-there-any-work-going-on-to-make-conversions-from-tensorflow-to-keras-and-vice-versa-possible) - 1 upvote - [How to I convert convert a pytorch model to keras model](https://www.g2.com/discussions/how-to-i-convert-convert-a-pytorch-model-to-keras-model) - 1 upvote - [How can I use Keras for](https://www.g2.com/discussions/37002-how-can-i-use-keras-for) - [View Keras pricing details and edition comparison](https://www.g2.com/products/keras/reviews?section=pricing&secure%5Bexpires_at%5D=2026-06-03+13%3A53%3A24+-0500&secure%5Bsession_id%5D=775d29f8-12ee-4ce5-8fa5-d14a0a0a5a98&secure%5Btoken%5D=46bf319542004092dcb86f5f114ffe4e0e3830bc5a1fe537eefb418dd3db7902&format=llm_user) ## Keras Integrations - [TensorFlow](https://www.g2.com/products/tensorflow/reviews) ## Keras Features **Core Functionality - Artificial Neural Network** - Neural Network Training - Neural Network Testing - Model Evaluation - Compliance **Data Handling - Artificial Neural Network** - Data Integration - Data Preprocessing **Performance - Artificial Neural Network** - Model Optimization - Scalability **Usability - Artificial Neural Network** - User Interface - Documentation & Support - Customizability **Advanced Features - Artificial Neural Network** - Deep Learning Capabilities - Transfer Learning - Real-Time Processing - Automated Model Tuning - Visualization Tools **Agentic AI - Artificial Neural Network** - Autonomous Task Execution - Multi-step Planning - Cross-system Integration - Adaptive Learning - Natural Language Interaction - Proactive Assistance - Decision Making ## Top Keras Alternatives - [TFLearn](https://www.g2.com/products/tflearn/reviews) - 4.0/5.0 (20 reviews) - [NVIDIA Deep Learning GPU Training System (DIGITS)](https://www.g2.com/products/nvidia-deep-learning-gpu-training-system-digits/reviews) - 4.5/5.0 (22 reviews) - [Microsoft Cognitive Toolkit (Formerly CNTK)](https://www.g2.com/products/microsoft-cognitive-toolkit-formerly-cntk/reviews) - 4.2/5.0 (22 reviews)