virtualstarsmith/awesome-oss-ai-alternatives

# Awesome Open-Source AI Alternatives **[中文](README.zh-CN.md)** ## Contents - [1. Cross-category Awesome Lists](#-1-cross-category-awesome-lists) - [2. ChatGPT / LLM Chat UI](#-2-chatgpt--llm-chat-ui) - [3. AI Design](#-3-ai-design) - [4. AI Coding](#-4-ai-coding) - [5. AI Image / Video / Audio / Avatar](#-5-ai-image--video--audio--avatar) - [6. Agents & Workflow Automation](#-6-agents--workflow-automation) - [7. AI Search](#-7-ai-search) - [8. Local Inference Engines](#-8-local-inference-engines) - [9. AI Slides](#-9-ai-slides) - [10. MCP Server Ecosystem](#-10-mcp-server-ecosystem) - [11. Browser Automation & Document OCR](#-11-browser-automation--document-ocr) - [12. AI Translation & Multilingual](#-12-ai-translation--multilingual) - [Quick Map](#quick-map) ## 1. Cross-category Awesome Lists Start here. These meta-lists are the fastest way to discover what tools exist in any AI sub-domain. - [**steven2358/awesome-generative-ai**](https://github.com/steven2358/awesome-generative-ai) — Encyclopedic generative-AI list across text / image / video / audio / code / agents. *(meta-list)* - [**alvinreal/awesome-opensource-ai**](https://github.com/alvinreal/awesome-opensource-ai) — "Truly open-source" AI projects only, no source-available pretenders. *(meta-list)* - [**caramaschiHG/awesome-ai-agents-2026**](https://github.com/caramaschiHG/awesome-ai-agents-2026) — 300+ resources mapping the 2026 agent landscape. *(meta-list)* - [**Anil-matcha/Open-Generative-AI**](https://github.com/Anil-matcha/Open-Generative-AI) — A direct "paid tool → open-source alternative" mapping table. *(meta-list)* - [**VoltAgent/awesome-agent-skills**](https://github.com/VoltAgent/awesome-agent-skills) — Plug-and-play Agent Skills you can drop into Claude / Cursor / Hermes. *(meta-list)* ## 2. ChatGPT / LLM Chat UI Self-hosted chat front-ends. The "private ChatGPT for my team / company / homelab" stack. - [**open-webui/open-webui**](https://github.com/open-webui/open-webui) — The de-facto self-hosted ChatGPT UI, 138K★. *(alt: ChatGPT)* - [**lobehub/lobe-chat**](https://github.com/lobehub/lobe-chat) — Beautifully branded multi-model chat, 77K★. *(alt: ChatGPT / Claude.ai)* - [**janhq/jan**](https://github.com/janhq/jan) — 100% offline ChatGPT desktop, runs llama.cpp locally. *(alt: ChatGPT Desktop)* - [**Mintplex-Labs/anything-llm**](https://github.com/Mintplex-Labs/anything-llm) — All-in-one RAG / document chat app, 60K★. *(alt: ChatGPT with files / NotebookLM)* - [**khoj-ai/khoj**](https://github.com/khoj-ai/khoj) — Personal AI "second brain" over your notes & files. *(alt: NotebookLM / Notion AI)* - [**danny-avila/LibreChat**](https://github.com/danny-avila/LibreChat) — Enhanced ChatGPT clone with multi-provider support, 25K★. *(alt: ChatGPT Plus)* - [**oobabooga/text-generation-webui**](https://github.com/oobabooga/text-generation-webui) — Gradio web UI for running LLMs locally with LoRA training. *(alt: ChatGPT)* ## 3. AI Design AI-native design tools and design-system skill packs that replace Figma + Magician + AI plugins. - [**nexu-io/open-design**](https://github.com/nexu-io) — Claude Design open-source alternative, 13 CLIs × 31 Skills × 72 design systems. *(alt: Claude Design / Figma AI)* - [**open-pencil/open-pencil**](https://github.com/open-pencil) — AI-native Figma alternative built on Tauri. *(alt: Figma)* - [**VoltAgent/awesome-claude-design**](https://github.com/VoltAgent/awesome-claude-design) — 68 production-grade `DESIGN.md` templates. *(skills pack)* - [**penpot/penpot**](https://github.com/penpot/penpot) — The first open-source design & prototyping platform. *(alt: Figma)* - [**excalidraw/excalidraw**](https://github.com/excalidraw/excalidraw) — Hand-drawn-style diagrams with AI text-to-diagram. *(alt: Miro / Whimsical)* ## 4. AI Coding Local or self-hosted coding agents and copilots: Cursor, Copilot, Codeium replacements. - [**TabbyML/tabby**](https://github.com/TabbyML/tabby) — Self-hosted GitHub Copilot, runs on your own GPU. *(alt: GitHub Copilot)* - [**continuedev/continue**](https://github.com/continuedev/continue) — Open-source IDE autopilot for VS Code & JetBrains. *(alt: Cursor / Copilot Chat)* - [**Aider-AI/aider**](https://github.com/Aider-AI/aider) — Terminal-native AI pair programmer with git-aware edits. *(alt: Cursor)* - [**cline/cline**](https://github.com/cline/cline) — Autonomous coding agent inside VS Code with file & terminal access. *(alt: Cursor Composer / Devin)* - [**OpenHands/OpenHands**](https://github.com/OpenHands/OpenHands) — Open-source software-engineer agent (formerly OpenDevin). *(alt: Devin)* - [**NousResearch/hermes-agent**](https://github.com/NousResearch) — Self-improving AI agent with FTS5 cross-session memory. *(alt: Claude Code / Cursor Composer)* - [**voideditor/void**](https://github.com/voideditor/void) — Open-source Cursor alternative with privacy-first design. *(alt: Cursor)* - [**trypear/pear-ai**](https://github.com/trypear/pearai-master) — Open-source AI code editor forked from VS Code. *(alt: Cursor)* ## 5. AI Image / Video / Audio / Avatar The "content creator's stack": image gen, video gen, music gen, voice clone, talking avatar — all open-source. - [**lllyasviel/Fooocus**](https://github.com/lllyasviel/Fooocus) — Midjourney-style UX on top of Stable Diffusion, 48K★. *(alt: Midjourney)* - [**Comfy-Org/ComfyUI**](https://github.com/Comfy-Org/ComfyUI) — Node-based Stable Diffusion / Flux pipeline editor. *(alt: Midjourney / RunwayML pipelines)* - [**invoke-ai/InvokeAI**](https://github.com/invoke-ai/InvokeAI) — Pro-grade Stable Diffusion creator suite. *(alt: Midjourney / Photoshop Generative Fill)* - [**Stability-AI/generative-models**](https://github.com/Stability-AI/generative-models) — Official SD3 / SVD video models. *(alt: Runway Gen-3)* - [**ace-step/ACE-Step-1.5**](https://github.com/ace-step) — Suno-grade music generation under 4GB VRAM. *(alt: Suno)* - [**multimodal-art-projection/YuE**](https://github.com/multimodal-art-projection/YuE) — Full-song generation foundation model. *(alt: Suno / Udio)* - [**RVC-Boss/GPT-SoVITS**](https://github.com/RVC-Boss/GPT-SoVITS) — 1-minute sample → high-fidelity voice clone. *(alt: ElevenLabs)* - [**coqui-ai/TTS**](https://github.com/coqui-ai/TTS) — Production-grade open TTS with 1100+ voices. *(alt: ElevenLabs / Play.ht)* - [**fishaudio/fish-speech**](https://github.com/fishaudio/fish-speech) — SOTA open-source TTS with accurate voice cloning from 10–30s samples, 30K★. *(alt: ElevenLabs)* - [**duixcom/Duix-Avatar**](https://github.com/duixcom) — Open-source talking avatar generator. *(alt: HeyGen / Synthesia)* - [**OpenTalker/SadTalker**](https://github.com/OpenTalker/SadTalker) — Single image + audio → talking head video. *(alt: HeyGen / D-ID)* - [**hpcaitech/Open-Sora**](https://github.com/hpcaitech/Open-Sora) — Open-source Sora alternative for high-quality video generation. *(alt: Sora / Runway Gen-3)* - [**AUTOMATIC1111/stable-diffusion-webui**](https://github.com/AUTOMATIC1111/stable-diffusion-webui) — The most popular Stable Diffusion web UI, 148K★. *(alt: Midjourney)* ## 6. Agents & Workflow Automation Visual & code-first automation backbones for content pipelines, hot-topic radars, multi-channel publishing. - [**n8n-io/n8n**](https://github.com/n8n-io/n8n) — Visual AI workflow builder, 189K★. *(alt: Zapier / Make)* - [**automatisch/automatisch**](https://github.com/automatisch/automatisch) — Pure open-source Zapier alternative, no enterprise lock-in. *(alt: Zapier)* - [**langflow-ai/langflow**](https://github.com/langflow-ai/langflow) — Drag-and-drop LangChain flows. *(alt: Flowise / Voiceflow AI)* - [**FlowiseAI/Flowise**](https://github.com/FlowiseAI/Flowise) — No-code LLM apps with chains, agents & RAG. *(alt: LangSmith / Voiceflow)* - [**Significant-Gravitas/AutoGPT**](https://github.com/Significant-Gravitas/AutoGPT) — Long-running autonomous agent platform. *(alt: agent.ai / AgentGPT cloud)* - [**crewAIInc/crewAI**](https://github.com/crewAIInc/crewAI) — Multi-agent role-based orchestration framework. *(alt: Multi-On)* - [**langgenius/dify**](https://github.com/langgenius/dify) — Production-ready LLM app development platform with workflow + RAG + agent, 130K★. *(alt: LangSmith / Voiceflow)* - [**activepieces/activepieces**](https://github.com/activepieces/activepieces) — Open-source Zapier alternative with AI pieces, 15K★. *(alt: Zapier)* ## 7. AI Search Self-hosted answer engines that combine web search + LLM with citations. - [**ItzCrazyKns/Vane**](https://github.com/ItzCrazyKns/Vane) — Privacy-first Perplexity alternative (formerly Perplexica), 34.5K★. *(alt: Perplexity)* - [**zaidmukaddam/scira**](https://github.com/zaidmukaddam/scira) — Multi-source AI search engine, 11.7K★. *(alt: Perplexity)* - [**nashsu/FreeAskInternet**](https://github.com/nashsu/FreeAskInternet) — 100% free, local LLM + multi-engine search. *(alt: Perplexity)* - [**YassKhazzan/openperplex_backend_os**](https://github.com/YassKhazzan/openperplex_backend_os) — Production-ready Perplexity backend in Python. *(alt: Perplexity API)* - [**infiniflow/ragflow**](https://github.com/infiniflow/ragflow) — Open-source RAG engine with citations and agent templates, 70K★. *(alt: Perplexity / ChatGPT Search)* ## 8. Local Inference Engines The runtimes that power everything above when you don't want to call OpenAI. - [**ollama/ollama**](https://github.com/ollama/ollama) — De-facto standard for running LLMs locally, 172K★. *(alt: OpenAI API for local)* - [**ggml-org/llama.cpp**](https://github.com/ggml-org/llama.cpp) — The C++ kernel behind most local-LLM tools. *(infra)* - [**vllm-project/vllm**](https://github.com/vllm-project/vllm) — High-throughput LLM serving with PagedAttention. *(alt: OpenAI / Together.ai serving)* - [**mudler/LocalAI**](https://github.com/mudler/LocalAI) — Drop-in OpenAI-compatible API for local models. *(alt: OpenAI API)* - [**mozilla/llamafile**](https://github.com/mozilla/llamafile) — Distribute and run LLMs as a single file, by Mozilla. *(alt: OpenAI API for local)* - [**huggingface/text-generation-inference**](https://github.com/huggingface/text-generation-inference) — Production LLM serving with continuous batching, by HuggingFace. *(alt: OpenAI API serving)* ## 9. AI Slides Generate full decks from a prompt, markdown, or doc — open & self-hostable. - [**nexu-io/html-anything**](https://github.com/nexu-io/html-anything) — Agentic HTML editor with 75+ skill templates, 9 export formats (slides, posters, articles, WeChat, Xiaohongshu). *(alt: Gamma / Canva)* - [**presenton/presenton**](https://github.com/presenton/presenton) — Open-source AI slide generator, Gamma alternative. *(alt: Gamma / Beautiful.ai)* - [**slidevjs/slidev**](https://github.com/slidevjs/slidev) — Markdown-first slides for developers with AI add-ons. *(alt: Gamma)* ## 10. MCP Server Ecosystem The hottest 2026 ecosystem: Model Context Protocol servers that let agents talk to anything. - [**punkpeye/awesome-mcp-servers**](https://github.com/punkpeye/awesome-mcp-servers) — The flagship MCP server directory, 82.9K★. *(meta-list)* - [**modelcontextprotocol/servers**](https://github.com/modelcontextprotocol/servers) — Official MCP reference servers. *(infra)* - [**wong2/awesome-mcp-clients**](https://github.com/wong2/awesome-mcp-clients) — Directory of MCP clients (Claude, Cursor, Cline, …). *(meta-list)* ## 11. Browser Automation & Document OCR "Agent goes online" and "paid PDFs → LLM-ready markdown" — two must-have building blocks. - [**microsoft/playwright**](https://github.com/microsoft/playwright) — Browser automation + AI Test Agent, 89K★. *(infra)* - [**browser-use/browser-use**](https://github.com/browser-use/browser-use) — Browser-Agent-only abstraction on top of Playwright. *(alt: Browserbase agent SDK)* - [**Skyvern-AI/skyvern**](https://github.com/Skyvern-AI/skyvern) — LLM-powered RPA for messy real-world websites. *(alt: UiPath AI)* - [**opendatalab/MinerU**](https://github.com/opendatalab/MinerU) — PDF / Office → LLM-ready Markdown converter. *(alt: Mistral OCR / Reducto)* - [**Unstructured-IO/unstructured**](https://github.com/Unstructured-IO/unstructured) — ETL pipeline for unstructured docs → vectors. *(alt: LlamaParse / Mistral OCR)* - [**dsdanielpark/docling**](https://github.com/docling-project/docling) — Document conversion to structured output (PDF / PPTX / HTML → JSON/Markdown). *(alt: Mistral OCR / Reducto)* ## 12. AI Translation & Multilingual Self-hosted translation, on par with DeepL for many language pairs. - [**LibreTranslate/LibreTranslate**](https://github.com/LibreTranslate/LibreTranslate) — Free, self-hosted translation API. *(alt: DeepL / Google Translate)* - [**thedaviddelta/lingva-translate**](https://github.com/thedaviddelta/lingva-translate) — Privacy-preserving Google Translate frontend. *(alt: Google Translate)* - [**OpenNMT/CTranslate2**](https://github.com/OpenNMT/CTranslate2) — Fast inference engine for NMT models. *(infra)* - [**argosopentech/argos-translate**](https://github.com/argosopentech/argos-translate) — Open-source offline translation library powered by OpenNMT. *(alt: Google Translate / DeepL)* ## Quick Map | Commercial Product | Open-source Pick | Category | |---|---|---| | ChatGPT (UI) | open-webui, lobe-chat, jan, LibreChat | Chat UI | | Figma / Claude Design | open-pencil, open-design, penpot | Design | | Cursor / Copilot | continue, aider, cline, tabby, void, pear-ai | Coding | | Midjourney | Fooocus, ComfyUI, InvokeAI, sd-webui | Image | | Suno / Udio | ACE-Step, YuE | Music | | ElevenLabs | GPT-SoVITS, Coqui TTS, fish-speech | Voice | | HeyGen / Synthesia | Duix-Avatar, SadTalker | Avatar | | Zapier / Make | n8n, automatisch, dify, activepieces | Automation | | Perplexity | Perplexica, scira, RAGFlow | Search | | Gamma / Beautiful.ai | html-anything, presenton, slidev | Slides | | DeepL | LibreTranslate | Translate | | Mistral OCR / Reducto | MinerU, unstructured, docling | Doc Parsing | | OpenAI API (hosted) | ollama, vLLM, LocalAI, llamafile, TGI | Local Inference |