notorch
Neural networks in pure C are provided by notorch, a software library in this repository.
About notorch
notorch is a neural network framework written in pure C that runs models and trains them. It compiles with a C compiler and includes a small harness that takes a GGUF file and a text prompt, with terminal chat when no prompt is given. It supports specific model families through separate files behind one interface and refuses files from unclaimed architectures by name. Output comparison with llama.cpp is done at temperature 0 and requires identical tokenization.
Key features
- Runs GGUF models from a single compiled binary
- Trains neural networks directly in pure C
- Supports llama, mistral3, qwen2, qwen3, gemma3, gemma4, MoE and Mamba
- Terminal chat mode when run without a prompt
- Compares output bytes against llama.cpp at temperature 0
- Refuses unclaimed architectures by name instead of defaulting
Best for: Developers who work in C and want to run and train neural networks without Python.
Good for
- Builds with cc and math library in a couple seconds
- Single binary with no runtime for inference harness
- Checks output against llama.cpp on the same weights
Watch out for
- Harness refuses files from architectures it does not claim
- Reference check does not judge files with different tokenization
- Some architecture code had limited file testing at time of writing
notorch at a glance
- Category
- Development › Developer Tools
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 5, 2026 · Online when checked
- Listed
- Oct 5, 2026
Guides for Developer Tools
notorch alternatives
Tools and sites you could use instead, ranked by how closely they match.
Efficient fine-tuning of quantized language models is the focus of QLoRA, a method documented in this repository.
Open source
24 alternatives
Neofuzz provides Python developers with a library for fuzzy text search.
Open source
24 alternatives
Kinho’s Neural repository provides an API for deep learning and feedforward neural network models.
Open source
24 alternatives
Gitingest turns a Git repository into a plain text digest designed for use with language model prompts.
crystal-fann provides Crystal language bindings for the FANN artificial neural network library.
Open source
24 alternatives
Netron displays neural network, deep learning, and machine learning models across formats such as ONNX, TensorFlow, and PyTorch.
Open source
24 alternativesPopular in Developer Tools
Well-known listings from the same topic.
AppDevToolsappdevtools.comCommand Not Foundcommand-not-found.comTestFlighttestflight.apple.com/join/A8GwCnq8Sourcegraphabout.sourcegraph.comProject Wallaceprojectwallace.comCarbon Calculatorwebsitecarbon.com
notorch FAQ
How do I run a model?
Build the harness with make harness, then run ./notorch model.gguf "prompt" with token count and temperature, or run without a prompt for terminal chat.
What is needed to build it?
C source compiled as cc notorch.c -o notorch -lm.
How is correctness checked?
make test_reference runs notorch and llama.cpp on the same weights at temperature 0 and compares the bytes.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.