Library
85,340 resources from 11 public sources. 54 checked by hand; the rest are unreviewed.
Results
-
BigMoeOnEdge
AI repositoriesRun MoE models bigger than your RAM. Frontier-size MoE on a phone, laptop, edge. Lossless, on stock llama.cpp. An adaptive streaming optimizer. Any hardware, any OS, any model, any quantization, behind a local OpenAI-compatible server.
git clone https://github.com/Helldez/BigMoeOnEdge.git -
whistle
Modelsautomatic speech recognition model by Cactus-Compute · cactus-compute
Cactus-Compute/whistle -
MasterAgent
Agents & frameworksBuild AI agents that run 100% on-device. Sub-100ms latency on Qualcomm NPU. Zero cloud dependency.
git clone https://github.com/OpenSparX/MasterAgent.git -
py-xiaozhi
MCP serversOpen-source AI assistant ecosystem with MCP integrations, multimodal workflows, IoT support, and cross-platform voice interaction.
git clone https://github.com/huangjunsen0406/py-xiaozhi.git -
MNN-Pocket-0.3B
Modelstext generation model by taobao-mnn · transformers
taobao-mnn/MNN-Pocket-0.3B -
LFM2.5-230M
Modelstext generation model by LiquidAI · transformers
LiquidAI/LFM2.5-230M -
LFM3-Omni-2B
Modelsany to any model by LiquidAI · transformers
LiquidAI/LFM3-Omni-2B -
Capri-AI-Desktop
Agents & frameworksCapri — Edgerunners-inspired AI agent for your desktop. Supports Claude 3.7, Claude Code, OpenAI, and local Ollama/LLaMA. Custom AI Waifu Companion with voice chat.
git clone https://github.com/PotterRemember37/Capri-AI-Desktop.git -
Audio8-TTS-Realistic-Air
Modelstext to speech model by Edge0 · transformers
Edge0/Audio8-TTS-Realistic-Air -
nodedb
AI repositoriesThe memory & storage database engine for AI agents. Multi-model, edge-to-cloud, PostgreSQL-compatible. Supports graph, vector, array, timeseries.
git clone https://github.com/NodeDB-Lab/nodedb.git -
d1-3B
Modelsimage text to text model by LiquidAI · transformers
LiquidAI/d1-3B -
kalshi-ai-trading-bot
Agents & frameworksThe honest edge layer for Kalshi prediction markets - an agent-native toolkit (CLI, MCP server, Claude skill) that measures whether you actually beat the book instead of claiming it.
git clone https://github.com/ryanfrigo/kalshi-ai-trading-bot.git -
Audio8-TTS-Preview-0.6b
Modelstext to speech model by Edge0 · transformers
Edge0/Audio8-TTS-Preview-0.6b -
LFM3-Audio-2B
Modelsany to any model by LiquidAI · transformers
LiquidAI/LFM3-Audio-2B -
LFM3-VL-2B
Modelsimage text to text model by LiquidAI · transformers
LiquidAI/LFM3-VL-2B -
kubesphere
SkillsThe container platform tailored for Kubernetes multi-cloud, datacenter, and edge management ⎈ 🖥 ☁️
git clone https://github.com/kubesphere/kubesphere.git -
iot-dc3
MCP serversIoT DC3 — Connect the Physical World to AI. An open-source Industrial IoT Runtime for Physical AI: 36 protocol drivers, edge-to-cloud data delivery, and an MCP tool gateway so AI agents act on devices in a safe, controllable, traceable loop.
git clone https://github.com/pnoker/iot-dc3.git -
rulego
Tools & apps⛓️RuleGo is a lightweight, high-performance, embedded, next-generation component orchestration rule engine framework for Go.
git clone https://github.com/rulego/rulego.git -
MNN
AI repositoriesMNN: A blazing-fast, lightweight inference engine battle-tested by Alibaba, powering high-performance on-device LLMs and Edge AI.
git clone https://github.com/alibaba/MNN.git -
LFM3-1.7B-Base
Modelstext generation model by LiquidAI · transformers
LiquidAI/LFM3-1.7B-Base -
d1-2B
Modelsimage text to text model by LiquidAI · transformers
LiquidAI/d1-2B -
daily
Tools & appsdaily.dev is the personalized developer news feed and community. Get the best tech content from all over the web in your browser new tab or on mobile. Free and open source.
git clone https://github.com/dailydotdev/daily.git -
LFM2.5-VL-1.6B
Modelsimage text to text model by LiquidAI · transformers
LiquidAI/LFM2.5-VL-1.6B -
dapr
Agents & frameworksDapr is a portable runtime for building distributed applications across cloud and edge, combining event-driven architecture with workflow orchestration.
git clone https://github.com/dapr/dapr.git
Indexed entries aren't reviewed or tested by InnerVault. Details come straight from each source. Anything that can reach your accounts, files or browser can do damage, so read the source before you install it.