TensorFlow is Google's open-source platform for building, training and deploying machine learning models, from phones to large clusters of machines.
In plain words
Machine learning code needs a lot of plumbing: math on large arrays, running on GPUs, training across many machines and serving models to users. TensorFlow bundles tools for all of it. It began inside the Google Brain team and was released as open source in 2015. It has stable Python and C++ interfaces, and Keras is its high-level way to build models.
A quick example
A typical workflow: load an image dataset, define a small model with Keras, train it on a GPU, then serve it with TensorFlow Serving so an app can send it pictures and get predictions back.
Where you'll meet it
- Training and serving machine learning models in production.
- Running models on mobile and edge devices.
Related
Sources
- Docstensorflow/tensorflow, TensorFlow (GitHub) · read 29 Sept 2026
- PaperTensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems, Abadi et al. (arXiv) · read 29 Sept 2026
- DocsLearn TensorFlow, TensorFlow · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.