An open-source alternative to TensorRT or CoreML for edge computing hardware. It optimizes deep learning model execution across heterogeneous computing resources like CPUs and NPUs.
Tech Stack
C++AI, ML & Data
Why Tengine?
• Edge optimized
• Modular design
• Multi-NPU support
Limitations
• Hardware specific
• Complex C++ API
• Limited model formats
3/5/2026
Last Update
979
Forks
249
Issues
Apache-2.0
License
Financial Leak Detected
Stop the "SaaS Tax"
Your team could be burning cash. Switching to Tengine instantly boosts your runway.
