Learn by Source
DeepLearning.AI · on Coursera

TensorFlow Developer Professional Certificate

4.7(28,000) on Coursera·430K enrolled
Intermediate 160 hours English Professional Certificate
SkillsTensorFlowDeep learningNeural networksComputer visionNLPTime series

Is this course right for you?

Our take
A hands-on, well-structured way to learn TensorFlow once you have Python and ML basics. It is free to audit.

Good for: learners with Python and ML basics who want hands-on TensorFlow skills.

Skip if: you are new to machine learning, or you specifically wanted the retired TF exam credential.

Laurence Moroney takes you through building neural networks with Keras, CNNs for image classification, transfer learning, and NLP and time-series models, with Colab exercises throughout. It is practical rather than theory-first.

One important caveat: Google retired the separate official TensorFlow Developer Certificate exam in 2024, so this is a course specialization, not a route to that credential. It also assumes ML basics, so it is not a first ML course — start with Ng's specialization if you are new. DeepLearning.AI keeps it updated.

Comparison · LBS

Compare alternatives for TensorFlow Developer Professional Certificate

Same topic, different options. We surface the trade-offs others hide so you can pick the course that actually fits your time, budget, and goals.
Coursera4.7(28,000)
TensorFlow Developer Professional Certificate
Price
Paid
Subscription-based, free to audit
Duration
160 hrs
Level
Intermediate
Certificate
Professional Certificate
Coursera
Introduction to TensorFlow for AI, ML and Deep Learning
Price
Free
Audit free · Certificate available
Duration
—
Level
Intermediate
Certificate
Professional Certificate
View course →
Udemy4.7(6,000)
Computer Vision with PyTorch: Deep Learning for Images
Price
Paid
Paid, frequently discounted
Duration
20 hrs
Level
Intermediate
Certificate
Course Certificate
View course →
Udemy4.6(18,000)
PyTorch for Deep Learning Bootcamp
Price
Paid
Paid, frequently discounted
Duration
17 hrs
Level
Intermediate
Certificate
Bootcamp
View course →
Prices & availability can change — confirm on the provider's site. We're not affiliated with any single provider.

About this course

This four-course specialization from DeepLearning.AI teaches you to build models in TensorFlow with Keras: neural network fundamentals, CNNs for image classification, transfer learning, NLP for text, and time-series forecasting with RNNs and LSTMs. Every course pairs short lessons with hands-on Colab exercises, so you learn by building rather than watching.

Browse all AI & ML courses Browse all Coursera courses

Instructor

LM
Laurence Moroney
Coursera instructor
430K+ learners5 courses4.7 instructor rating

Taught by Laurence Moroney, Google's AI Advocate who designed the TensorFlow Developer Certificate program.

Frequently asked questions

No — this is the single most important thing to know. Google closed the official exam in mid-2024 and has announced no replacement, so it can no longer be purchased or registered for. Certificates awarded before the closure remain valid for three years, but new candidates simply cannot earn this credential any more. Any course still selling it as a path to the certificate is out of date.

The skills are, even though the credential is not. Learning to build and train models in TensorFlow remains genuinely useful, and the material itself did not stop being valid when the exam closed. Just go in for the practical ability rather than a certificate you can no longer obtain, and set your expectations — and any spending — accordingly, treating it as a TensorFlow course rather than exam prep.

PyTorch has become the default in research and much of industry, and most new models appear there first, so if you are starting fresh it is the more strategic bet for job prospects today. TensorFlow is far from dead — it is still widely deployed in production systems — so learning it is not wasted, but weigh which one the roles or projects you are targeting actually use before committing your time.

Comfortable Python and a working grasp of core machine-learning ideas such as training, loss, and overfitting. It also helps to have seen neural networks before, because the material focuses on implementing them in TensorFlow rather than teaching the underlying theory from scratch. Arriving without those foundations means learning two hard things at once, which is a tough way to approach a framework-specific course.

Building neural networks with TensorFlow and Keras across the main problem types: image classification with convolutional networks, natural language processing, and time-series or sequence data. The emphasis is hands-on model building — writing and training real networks — rather than deep mathematical derivation, which makes it practical, though it does assume you already understand the concepts those models implement.

The training materials can usually be audited or accessed free depending on the platform, though graded elements and any completion certificate may need payment. Because the official Google exam no longer exists, there is little point spending money chasing a credential — if you do pay for anything, put it toward hands-on practice and projects rather than a certificate that no longer carries the weight it once did.
Paid
Subscription-based, free to audit
Enroll now