Library/MNNForked

alibaba/MNN

MNN

MNN: A blazing-fast, lightweight inference engine battle-tested by Alibaba, powering high-performance on-device LLMs and Edge AI.

Builder

alibaba

alibaba

alibaba • individual

Stars

14,745

Using upstream star count

Forks

2,268

Using upstream fork count

Open Issues

0

Activity Score

0/100

0 commits in 30d

Created

Apr 15, 2019

Project creation date

README Summary

MNN is a high-performance, lightweight deep learning inference engine developed by Alibaba that is optimized for mobile and edge devices. It supports multiple platforms and provides efficient execution of neural networks with minimal resource consumption. The framework is battle-tested in production environments and designed specifically for on-device AI applications including large language models.

AI Dev Skills

Unmapped

Deep Learning Inference OptimizationCross-platform Mobile AIEdge ComputingNeural Network QuantizationHardware AccelerationModel CompressionONNX Model ConversionTensorFlow Lite IntegrationPyTorch Model DeploymentComputer Vision InferenceNatural Language Processing InferenceGPU ComputingCPU Optimization

Tags

Deep Learning Inference OptimizationCross-platform Mobile AIEdge ComputingNeural Network QuantizationHardware AccelerationModel CompressionONNX Model ConversionTensorFlow Lite IntegrationPyTorch Model DeploymentComputer Vision InferenceNatural Language Processing InferenceGPU ComputingCPU OptimizationTextSmart Camera ProcessingImageMobile AI ApplicationsVideoOn-device AISmart DevicesEdge AI InferenceIoTEmbedded SystemsMobile ApplicationsAutonomous Vehicle PerceptionTelecommunicationsMultimodalModel OptimizationAudioRoboticsAutomotiveOn-device Natural Language ProcessingMobile AIVoice Assistant DeploymentEfficient AIIoT Device IntelligenceReal-time Image RecognitionEdge AIEdge/MobileOn-premiseC++

Taxonomy

Recent Activity

Updated 24 days ago

7 Days

0

30 Days

0

90 Days

0

Quality

production
Quality
high
Maturity
production

Categories

Inference & ServingPrimaryDev Tools & AutomationNLP & TextML Platform & InfrastructureCoding & Dev ToolsMultimodal AIEdge & Mobile AIOther AI / MLRoboticsFoundation ModelsAI AgentsModel TrainingGenerative MediaComputer Vision

PM Skills

Scale & Reliability

Languages

C++100.0%

Timeline

Project created
Apr 15, 2019
Forked
Mar 22, 2026
Your last push
24 days ago
Upstream last push
6 days ago
Tracked since
Mar 20, 2026

Similar Repos

pgvector cosine similarity · $0

Loading…