MindSpore provides a deep learning framework for training and inference across mobile, edge, and cloud scenarios.
Open source
21 alternativesTNN
Uniform deep learning inference framework for mobile, desktop and server.
About TNN
TNN, developed by Tencent Youtu Lab and Guangying Lab, is a neural network inference framework for mobile, desktop and server use. Deployment follows three steps: convert a trained model to TNN format, compile the engine for the target hardware, and call the engine from an application for inference.
It describes conversion support for TensorFlow, PyTorch and Caffe models, and engine builds for ARM, OpenCL, Metal, NPU, X86 and CUDA. The repository includes examples for face detection, face alignment, hair segmentation, pose estimation, object detection and Chinese OCR.
Key features
- Converts TensorFlow, PyTorch, and Caffe models to TNN format
- Builds target-specific engines for ARM, OpenCL, Metal, NPU, X86, CUDA
- Includes face detection example using BlazeFace
- Includes Chinese OCR example for tilted, rotated, and vertical text
- Includes object detection examples for YOLOv5s and MobileNetV2-SSD
- Supports model compression and code pruning
Best for: Developers deploying trained models for inference on mobile, desktop, or server hardware.
Good for
- Covers mobile, desktop, and server deployment targets
- Supports multiple hardware backends including ARM, NPU, X86, and CUDA
- Provides conversion tools and examples for vision and OCR tasks
TNN at a glance
- Website
- github.com/Tencent/TNN
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Open source
- Not stated
- Language
- English
- Last checked
- Oct 5, 2026 · Online when checked
- Listed
- Oct 5, 2026
TNN alternatives
Tools and sites you could use instead, ranked by how closely they match.
NVIDIA Triton Inference Server provides an inference solution for deploying models in cloud and edge environments.
Open source
23 alternatives
For Apple Silicon, Rapid-MLX offers an open-source LLM inference server and Mac app compatible with OpenAI and Anthropic APIs.
BitNet provides an inference framework for one-bit large language models.
Open source
23 alternatives
GraphPipe provides a protocol and software for deploying machine learning models separately from their frameworks.
Popular in Model Serving & Inference
Well-known listings from the same topic.
MAI Playgroundplayground.microsoft.ai/chatBigModelbigmodel.cnkimi-k2.6platform.kimi.ai/docs/pricing/chat-k26Jinagithub.com/jina-ai/jinaFluxgithub.com/black-forest-labs/fluxprivateGPTgithub.com/imartinez/privateGPT
TNN FAQ
What steps are needed to use TNN with a trained model?
Convert the trained model to a TNN model, compile the TNN engine for the target platform, then use the compiled engine for inference.
Which training frameworks are mentioned for model conversion?
TensorFlow, PyTorch, and Caffe are listed as supported for conversion to TNN format.
Which acceleration options are mentioned for compiling TNN?
ARM, OpenCL, Metal, NPU, X86, and CUDA are listed as selectable options.
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.