1
ollama / ollama
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
View ollama/ollamaTrending open-source projects · updated daily
Local & On-Device AI Projects
Running AI locally means the model, the weights and the data all stay on hardware you control: no API key, no per-token bill, no prompt leaving the machine. This board collects the projects that make that practical — local model runners and servers, chat interfaces and desktop assistants, mobile and on-device inference, and the quantization and compression tooling that lets a model fit into laptop or phone memory. Entries come from GitHub topic pages for local LLMs, llama.cpp-style inference, on-device AI and Ollama, ranked by stars, so a mature server sits next to a desktop client published last month. Each card lists language, stars and forks and opens a page with description, license, activity dates, README and related local AI projects. The practical question is almost always memory: check the quantization notes on the detail page before assuming a model will fit.
Trending Local & On-Device AI Projects
2
open-webui / open-webui
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
View open-webui/open-webui3
ChatGPTNextWeb / NextChat
✨ Zero-config AI chat assistant. No API key needed — sign up and instantly chat with GPT-5, Claude 4, Gemini 2.5, DeepSeek & 100+ top models. Pay-as-you-go saves you more.
View ChatGPTNextWeb/NextChat4
HKUDS / nanobot
Ultra-lightweight, open-source, self-hosted personal AI agent framework in Python with WebUI, tools, memory, MCP, multi-agent workflows, automation, and chat apps
View HKUDS/nanobot5
→6
chatchat-space / Langchain-Chatchat
Langchain-Chatchat（原Langchain-ChatGLM）基于 Langchain 与 ChatGLM, Qwen 与 Llama 等语言模型的 RAG 与 Agent 应用 | Langchain-Chatchat (formerly langchain-ChatGLM), local knowledge based LLM (like ChatGLM
View chatchat-space/Langchain-Chatchat7
1Panel-dev / 1Panel
🔥 1Panel is a modern, open-source Linux server management panel and a lightweight AI management platform.
View 1Panel-dev/1Panel8
Zackriya-Solutions / meetily
Privacy first, AI meeting assistant with 4x faster Parakeet/Whisper live transcription, speaker diarization, and Ollama summarization built on Rust. 100% local processing. no cloud required.
View Zackriya-Solutions/meetily9
Tencent / WeKnora
Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.
View Tencent/WeKnora10
→11
1Panel-dev / MaxKB
🔥 MaxKB is an open-source platform for building enterprise-grade agents. 强大易用的开源企业级智能体平台。
View 1Panel-dev/MaxKB12
dyad-sh / dyad
Local, open-source AI app builder for power users ✨ v0 / Lovable / Replit / Bolt alternative 🌟 Star if you like it!
View dyad-sh/dyad13
lss233 / kirara-ai
🤖 可 DIY 的 多模态 AI 聊天机器人 | 🚀 快速接入 微信、 QQ、Telegram、等聊天平台 | 🦈支持DeepSeek、Grok、Claude、Ollama、Gemini、OpenAI | 工作流系统、网页搜索、AI画图、人设调教、虚拟女仆、语音对话 |
View lss233/kirara-ai14
AsyncFuncAI / deepwiki-open
Open Source DeepWiki: AI-Powered Wiki Generator for GitHub/Gitlab/Bitbucket Repositories. Join the discord: https://discord.gg/gMwThUMeme
View AsyncFuncAI/deepwiki-open15
langbot-app / LangBot
Production-grade platform for building agentic IM bots - 生产级多平台智能机器人开发平台/ Agent、知识库编排、插件系统 / Bots for Discord / Slack / LINE / Telegram / WeChat(企业微信, 企微智能机器人, 公众号) / 飞书 / 钉钉 / QQ / Matrix e.g.
View langbot-app/LangBot16
MODSetter / SurfSense
Air gapped, open source NotebookLM alternative. Join our Discord: https://discord.gg/ejRNvftDp9
View MODSetter/SurfSense17
budtmo / docker-android
Android in docker solution with noVNC supported, video recording, mcp server and AI-agent
View budtmo/docker-android18
lidge-jun / opencodex
Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama…) with Codex CLI, App, SDK, and Claude Code
View lidge-jun/opencodex19
GaiZhenbiao / ChuanhuChatGPT
GUI for ChatGPT API and many LLMs. Supports agents, file-based QA, GPT finetuning and query with web search. All with a neat UI.
View GaiZhenbiao/ChuanhuChatGPT20
tonhowtf / omniget
Udemy & Hotmart course downloader, YouTube downloader (yt-dlp GUI, 1,800+ sites) + desktop app for AI agents: Claude Code, Codex, Gemini CLI, Ollama.
View tonhowtf/omniget21
Open-LLM-VTuber / Open-LLM-VTuber
Talk to any LLM with hands-free voice interaction, voice interruption, and Live2D avatar running locally across platforms
View Open-LLM-VTuber/Open-LLM-VTuber22
browseros-ai / BrowserOS
🌐 The open-source Agentic browser; alternative to ChatGPT Atlas, Perplexity Comet, Dia.
View browseros-ai/BrowserOS23
langchain4j / langchain4j
LangChain4j is an idiomatic, open-source Java library for building LLM-powered applications on the JVM.
View langchain4j/langchain4j24
The-PR-Agent / pr-agent
🚀 PR Agent: The Original Open-Source PR Reviewer. This project is not the Qodo free tier.
View The-PR-Agent/pr-agent25
StarTrail-org / LEANN
[MLsys2026 Best Paper]: https://arxiv.org/abs/2506.08276. RAG on Everything with LEANN. Enjoy 97% storage savings while running a fast, accurate
View StarTrail-org/LEANN26
TheR1D / shell_gpt
A command-line productivity tool powered by AI large language models like GPT-5, will help you accomplish your tasks faster and more efficiently.
View TheR1D/shell_gpt27
cactus-compute / needle
Automation foundation model for tiny devices: 2-bit, 8-29 MB, tool calls, structured extraction and embeddings on phones, wearables, smart homes, robots, cars and microcontrollers.
View cactus-compute/needle28
altic-dev / FluidVoice
Fastest and only macOS Dictation app with on-device STT and custom trained AI enhancement model. Windows pre-build available! A local Wispr Flow alternative.
View altic-dev/FluidVoice29
getumbrel / llama-gpt
A self-hosted, offline, ChatGPT-like chatbot. Powered by Llama 2. 100% private, with no data leaving your device. New: Code Llama support!
View getumbrel/llama-gpt30
→31
sigoden / aichat
All-in-one LLM CLI tool featuring Shell Assistant, Chat-REPL, RAG, AI Tools & Agents, with access to OpenAI, Claude, Gemini, Ollama, Groq, and more.
View sigoden/aichat32
RunanywhereAI / runanywhere-sdks
Production ready toolkit to run AI locally
View RunanywhereAI/runanywhere-sdks33
xorbitsai / inference
Swap GPT for any LLM by changing a single line of code. Xinference lets you run open-source, speech, and multimodal models on cloud, on-prem, or your laptop — all through one unified
View xorbitsai/inference34
QwenAudio / SenseVoice
Open-source SenseVoiceSmall model for Mandarin, Cantonese, English, Japanese, and Korean ASR, language ID, emotion recognition, and audio event detection.
View QwenAudio/SenseVoice35
→36
LearningCircuit / local-deep-research
~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private documents.
View LearningCircuit/local-deep-research37
reorproject / reor
Private & local AI personal knowledge management app for high entropy people.
View reorproject/reor38
qualcomm / GenieX
Run frontier LLMs and VLMs locally on Qualcomm devices across NPU, GPU, and CPU with a few lines of code
View qualcomm/GenieX39
n4ze3m / page-assist
Use your locally running AI models to assist you in your web browsing
View n4ze3m/page-assist40
OpenCoworkAI / open-codesign
Open-source Claude Design alternative. One-click import your Claude Code / Codex API key. Prompt → prototype / slides / PDF. Multi-model (Claude, GPT, Gemini, Kimi, GLM, Ollama).
View OpenCoworkAI/open-codesign41
mnfst / awesome-free-llm-apis
List of Permanent Free LLM API (API Keys)
View mnfst/awesome-free-llm-apis42
ArvinLovegood / go-stock
🦄🦄🦄AI赋能股票分析：AI加持的股票分析/选股工具。股票行情获取，AI热点资讯分析，AI资金/财务分析，涨跌报警推送。支持A股，港股，美股。支持市场整体/个股情绪分析，AI辅助选股等。数据全部保留在本地。支持DeepSeek，OpenAI， Ollama，LMStudio，AnythingLLM，硅基流动，火山方舟，阿里云百炼等平台或模型。
View ArvinLovegood/go-stock43
JerryZLiu / Dayflow
The automatic work journal/time tracker. Privately turns your screen into a timeline of what you actually accomplished. Open-source and local-first.
View JerryZLiu/Dayflow44
open-multi-agent / open-multi-agent
Self-hosted TypeScript agent runtime with durable approvals and verifiable run records. Own it, approve it, audit it.
View open-multi-agent/open-multi-agent45
MakazhanAlpamys / Soup
Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.
View MakazhanAlpamys/Soup46
→47
drumih / turbo-fieldfare
Gemma 4 26B-A4B inference in ~2 GB of RAM on any M-series MacBook
View drumih/turbo-fieldfare48
vas3k / TaxHacker
Self-hosted AI accounting app. LLM analyzer for receipts, invoices, transactions with custom prompts and categories
View vas3k/TaxHacker49
Osmantic / ODS
Turn your PC, Mac, or Linux box into an AI server. LLM inference, chat UI, voice, agents, workflows, RAG, and image generation.
View Osmantic/ODS50
google-ai-edge / LiteRT-LM
LiteRT-LM is Google's production-ready, high-performance, open-source inference framework for deploying Large Language Models on edge devices.
View google-ai-edge/LiteRT-LM51
cactus-compute / cactus
Quantization, kernels, runtime and inference engine for mobiles, wearables, smart home and robots.
View cactus-compute/cactus52
gluonfield / enchanted
Enchanted is iOS and macOS app for chatting with private self hosted language models such as Llama2, Mistral or Vicuna using Ollama.
View gluonfield/enchanted53
clusterzx / paperless-ai
An automated document analyzer for Paperless-ngx using OpenAI API, Ollama, Deepseek-r1, Azure and all OpenAI API compatible Services to automatically analyze and tag your documents.
View clusterzx/paperless-ai54
AgentOps-AI / agentops
Python SDK for AI agent monitoring, LLM cost tracking, benchmarking, and more. Integrates with most LLMs and agent frameworks including CrewAI, Agno, OpenAI Agents SDK, Langchain, Autogen, AG2
View AgentOps-AI/agentops55
PySpur-Dev / pyspur
A visual playground for agentic workflows: Iterate over your agents 10x faster
View PySpur-Dev/pyspur56
dograh-hq / dograh
Open source voice AI platform. Self-hosted alternative to Vapi and Retell. On Prem, BYOK across Speech to Speech or LLM/STT/TTS, with a visual workflow builder, MCP native and telephony support.
View dograh-hq/dograh57
maziyarpanahi / openmed
Local-first healthcare AI: clinical NER & HIPAA PII de-identification that runs 100% on-device. 2,200+ medical models, 21 languages, Apple MLX + Python, no cloud
View maziyarpanahi/openmed58
buxuku / SmartSub
视频转字幕、字幕翻译、AI 配音与声音克隆、字幕烧录——免费开源的一站式桌面工具。基于 Whisper / FunASR 等本地模型离线语音转文字，批量处理 + 全平台 GPU 加速，跨 Windows / macOS / Linux。Free, open-source desktop app to generate, translate
View buxuku/SmartSub59
Kiln-AI / Kiln
Build, Evaluate, and Optimize AI Systems. Includes evals, RAG, agents, fine-tuning, synthetic data generation, dataset management, MCP, and more.
View Kiln-AI/Kiln60
vinta / pangu.js
Opinionated paranoid text spacing in JavaScript, with on-device AI semantic judgment
View vinta/pangu.js61
thunderbird / thunderbolt
AI You Control: Choose your models. Own your data. Eliminate vendor lock-in.
View thunderbird/thunderbolt62
u14app / deep-research
Use any LLMs (Large Language Models) for Deep Research. Support SSE API and MCP server.
View u14app/deep-research63
JetBrains / koog
Koog is a JVM (Java and Kotlin) framework for building predictable, fault-tolerant and enterprise-ready AI agents across all platforms – from backend services to Android and iOS, JVM
View JetBrains/koog64
CaviraOSS / LongMemory
Local persistent memory store for LLM applications including claude desktop, github copilot, codex, antigravity, etc.
View CaviraOSS/LongMemory65
crmne / ruby_llm
The Ruby-native AI framework. Chats, agents, tools, images, audio, and video through one consistent API, in plain Ruby or Rails.
View crmne/ruby_llm66
→67
umlx5h / LLPlayer
The media player for language learning, with dual subtitles, AI-generated subtitles, real-time translation, and more!
View umlx5h/LLPlayer68
GiovanniPasq / agentic-rag-for-dummies
A modular Agentic RAG built with LangGraph — learn Retrieval-Augmented Generation Agents in minutes.
View GiovanniPasq/agentic-rag-for-dummies69
→70
nyldn / claude-octopus
Run multiple AI models against the same research, design, or coding task. Surface disagreements before you ship.
View nyldn/claude-octopus71
yuruotong1 / autoMate
Like Manus, Computer Use Agent(CUA) and Omniparser, we are computer-using agents.AI-driven local automation assistant that uses natural language to make computers work by themselves
View yuruotong1/autoMate72
lemony-ai / cascadeflow
Cascading runtime for AI agents. Optimize cost, latency, quality, and policy decisions inside the agent loop.
View lemony-ai/cascadeflow73
claraverse-space / ClaraVerse
Claraverse is a opesource privacy focused ecosystem to replace ChatGPT, Claude, N8N, ImageGen with your own hosted llm, keys and compute. With desktop, IOS, Android Apps.
View claraverse-space/ClaraVerse74
SciSharp / LLamaSharp
A C#/.NET library to run LLM (🦙LLaMA/LLaVA) on your local device efficiently.
View SciSharp/LLamaSharp75
raullenchai / Rapid-MLX
The fastest local AI engine for Apple Silicon. 4.2x faster than Ollama, 0.08s cached TTFT, 100% tool calling. 17 tool parsers, prompt cache, reasoning separation, cloud routing.
View raullenchai/Rapid-MLX76
→77
pipeshub-ai / pipeshub-ai
PipesHub is an open-source platform for securely connecting enterprise knowledge to AI. Give AI agents trusted context and your team permission-aware search with verified citations across your
View pipeshub-ai/pipeshub-ai78
OvidijusParsiunas / deep-chat
Fully customizable AI chatbot component for your website
View OvidijusParsiunas/deep-chat79
twinnydotdev / twinny
The most no-nonsense, locally or API-hosted AI code completion plugin for Visual Studio Code - like GitHub Copilot but 100% free.
View twinnydotdev/twinny80
deta / surf
Personal AI Notebooks. Organize files & webpages and generate notes from them. Open source, local & open data, open model choice (incl. local).
View deta/surf81
→82
→83
QiuYannnn / Local-File-Organizer
An AI-powered file management tool that ensures privacy by organizing local texts, images. Using Llama3.2 3B and Llava v1.6 models with the Nexa SDK, it intuitively scans, restructures
View QiuYannnn/Local-File-Organizer84
nicedreamzapp / claude-code-local
Run Claude Code 100% on-device with local AI on Apple Silicon. MLX-native Anthropic-API server. 6 fighters incl.
View nicedreamzapp/claude-code-local85
rag-web-ui / rag-web-ui
RAG Web UI is an intelligent dialogue system based on RAG (Retrieval-Augmented Generation) technology.
View rag-web-ui/rag-web-ui86
off-grid-ai / OGAM
The Swiss Army Knife of Offline AI. Chat, see, speak, and generate images on your phone or Mac — GGUF LLMs, vision, Whisper speech-to-text, Stable Diffusion, tool calling, and local-network servers.
View off-grid-ai/OGAM87
SharpAI / DeepCamera
Open-Source AI Camera Skills Platform, AI NVR & CCTV Surveillance. Local VLM video analysis with Qwen, DeepSeek, SmolVLM, LLaVA, YOLO26.
View SharpAI/DeepCamera88
xyproto / algernon
Small self-contained pure-Go web server with Lua, Teal, Markdown, HTTP/2, QUIC, Redis, TypeScript, npm-less React 19, SQLite, and PostgreSQL support ++
View xyproto/algernon89
sherlockchou86 / VideoPipe
A cross-platform video structuring (video analysis) framework based on CV models & mLLM.
View sherlockchou86/VideoPipe90
→91
→92
rafska / awesome-local-llm
A curated list of awesome platforms, tools, practices and resources that helps run LLMs locally
View rafska/awesome-local-llm93
Luce-Org / lucebox
LLM speculative inference server for heterogeneous hardware & consumer GPUs
View Luce-Org/lucebox94
Mininglamp-AI / Mano-P
Mano-P: Open-source GUI-VLA agent for edge devices. #1 on OSWorld (specialized, 58.2%). Runs locally on Apple M4 Mac mini/MacBook
View Mininglamp-AI/Mano-P95
→96
icereed / paperless-gpt
Use LLMs and LLM Vision (OCR) to handle paperless-ngx - Document Digitalization powered by AI
View icereed/paperless-gpt97
Mobile-Artificial-Intelligence / maid
Maid is a free and open source application for interfacing with llama.cpp models locally, and with Anthropic, DeepSeek, Ollama, Mistral and OpenAI models remotely.
View Mobile-Artificial-Intelligence/maid98
nicepkg / aide
Conquer Any Code in VSCode: One-Click Comments, Conversions, UI-to-Code, and AI Batch Processing of Files! 在 VSCode 中征服任何代码：一键注释、转换、UI 图生成代码、AI 批量处理文件！💪
View nicepkg/aide99
codedogQBY / ReadAny
AI-powered cross-platform e-book reader with semantic search, RAG chat, local vector store, notes, TTS, and WebDAV sync.
View codedogQBY/ReadAny100
→Local Model Runners & Servers
- ollama/ollama — ★ 181,296
- open-webui/open-webui — ★ 152,601
- 1Panel-dev/1Panel — ★ 36,983
- Zackriya-Solutions/meetily — ★ 30,962
- lss233/kirara-ai — ★ 19,027
- langbot-app/LangBot — ★ 17,926
- budtmo/docker-android — ★ 15,874
- lidge-jun/opencodex — ★ 15,561
Chat UIs & Desktop Assistants
- ChatGPTNextWeb/NextChat — ★ 88,790
- HKUDS/nanobot — ★ 48,395
- chatboxai/chatbox — ★ 41,812
- chatchat-space/Langchain-Chatchat — ★ 38,648
- 1Panel-dev/MaxKB — ★ 22,844
- dyad-sh/dyad — ★ 21,587
- GaiZhenbiao/ChuanhuChatGPT — ★ 15,272
- tonhowtf/omniget — ★ 13,986
On-Device & Mobile AI
- Tencent/WeKnora — ★ 27,811
- altic-dev/FluidVoice — ★ 11,660
- reorproject/reor — ★ 8,552
- google-ai-edge/LiteRT-LM — ★ 6,482
- cactus-compute/cactus — ★ 6,033
- gluonfield/enchanted — ★ 6,002
- maziyarpanahi/openmed — ★ 5,346
- vinta/pangu.js — ★ 4,827
Quantization & Model Compression
- off-grid-ai/OGAM — ★ 3,129
- mozilla-ai/llamafile — ★ 26,001
- AsyncFuncAI/deepwiki-open — ★ 18,018
More AI Trending Projects
1
→2
→3
→4
→5
→6
→Running an LLM locally: what you actually need
A local model needs three things: weights in a format the runtime understands, enough memory to hold them, and a runner to load them. The rule of thumb for four-bit quantization is roughly two thirds of a gigabyte per billion parameters, plus room for the context window — so a 7B model fits comfortably in 8 GB of unified memory, while a 70B model does not fit on a laptop at all without aggressive compression or offloading into system RAM. Server-grade engines change that arithmetic with paging, continuous batching and GPU offload, which is why the same weights can be impractical in one tool and comfortable in another.
Start with a runner rather than a chat client. Once the model answers from the command line, everything above it is interface.
Local chat UIs, desktop assistants and on-device inference
The desktop layer is what most people actually want: a chat window that keeps transcripts on disk, can read local files, and talks to whichever local server is running. Those clients are deliberately model-agnostic — they speak the same HTTP API as the popular runners — so the choice of client and the choice of model are independent decisions. On phones the constraints change completely: memory is small, thermals are real, and the interesting projects are the ones shipping compressed models that run in a few hundred megabytes with acceptable latency, rather than the ones streaming a full-size model over the network.
Quantization: why the same model runs on your laptop
Quantization stores weights at lower precision — four bits instead of sixteen — which cuts memory use roughly fourfold and speeds up inference on hardware with no tensor cores. It is not free: below four bits the quality loss becomes visible on reasoning tasks, and some quantized formats only load in specific runtimes. The practical guidance is boring and effective: prefer four-bit formats for local use, keep a higher-precision copy on a server if you need to compare, and test the quantized build on your own inputs instead of trusting a benchmark table.
Local AI questions
What is a local LLM? A language model whose weights run on your own hardware, so prompts and outputs never leave the machine and there is no per-token cost.
Can a local chatbot work offline? Yes, once the model has been downloaded. Runners and desktop clients on this board keep working with the network unplugged; only features that fetch from the internet stop.
What is on-device AI? Inference that happens on the device that produced the data — phone, laptop or embedded board — rather than in a hosted service. Its binding constraints are memory bandwidth and thermal budget.
How much RAM do I need to run a model locally? Roughly two thirds of a gigabyte per billion parameters at four-bit quantization, plus context. An 8B model is comfortable in 8 GB; anything past 30B wants a workstation or a GPU with real VRAM.
Is a local model as good as a hosted one? For summarising, drafting and light coding a good 8B–14B model is often enough. For hard multi-step reasoning the largest hosted models still lead, and they lead by more the longer the task runs.