chainer/chainer

Health v2 · maintenance only

Adoption not part of the score

5924 stars · 1342 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Chainer is a Python-based deep learning framework that pioneered the define-by-run approach with dynamic computational graphs and automatic differentiation. It supports CUDA/cuDNN acceleration via CuPy and is now in a maintenance phase, receiving only bug fixes.

Use cases

• build and train neural networks in python
• train deep learning models on gpu with cuda
• define dynamic computational graphs with control flow
• run recurrent and recursive network architectures
• prototype deep learning research with define-by-run
• multi-gpu training of neural networks

When to choose

• you need a flexible define-by-run framework with dynamic graphs
• you are maintaining an existing Chainer codebase
• you want CUDA-accelerated training with minimal code changes

When to avoid

• you are starting a new deep learning project - PyTorch or TensorFlow are actively developed
• you need long-term community support and new features
• you need production-grade deployment tooling

Facets

framework · maturity maintenance

deep-learning machine-learning gpu-computing deep-learning machine-learning gpu-computing python define-by-run dynamic-computational-graphs cuda cupy automatic-differentiation neural-networks linux gpu docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="chainer/chainer")