ArchestackBest of Open-Source AI: List on GitHub ↗ (opens in a new tab)

Workflow automation with AI

Workflow and integration platforms with AI steps or agent nodes, for wiring models into the rest of your tools.

n8n leads with 95, ahead of Activepieces (82) and banana-slides (69). 3 projects ranked by score. Updated from public GitHub data.

Same topic: AI editors and workflow canvases (4, starters to fork). All categories

Coming from Zapier? See the open-source alternatives.

The ranking

Workflow automation with AI: full ranking
RankProjectAdoptionFreshnessMaintenanceAgent-readyScore
1n8nVisual workflow automation with AI agent nodes and 1500+ integrations208.5k starscustom licenselast commit Oct 2026100100927095 out of 100
2ActivepiecesSelf-hosted workflow automation with TypeScript integrations, alternative to Zapier25.2k starscustom licenselast commit Oct 202660100868582 out of 100
3banana-slidesAI slide deck generator that exports PPTX, PDF and narrated video Live demo ↗ (opens in a new tab)15.9k starsAGPL-3.0last commit Oct 20262794993069 out of 100

Momentum, Verified build, Docs and Privacy are not measured yet; their weight goes to the signals shown. A dash means the signal is not scored for that kind of project. Hover a number for its rating in words.

Reviews

195 out of 100

n8n

Visual workflow automation with AI agent nodes and 1500+ integrations

208.5k starscustom licenselast commit Oct 2026

n8n is a node-based canvas for building workflows and multi-step AI agents, with JavaScript, Python and npm packages usable inside nodes. It connects to OpenAI, Anthropic, Google and open-source models, and lists 1500+ integrations and 9,000+ workflow templates. It runs self-hosted or on n8n Cloud, under the Sustainable Use License rather than an OSI license.

Strengths

  • 1500+ integrations and 9,000+ templates cover most common SaaS wiring
  • Visual canvas with JavaScript, Python and npm code when needed
  • Model-agnostic: OpenAI, Anthropic, Google or open-source models
  • Supports human approvals and observability for multi-step AI workflows

Weaknesses

  • Fair-code Sustainable Use License, not OSI open source; use is restricted
  • Role-based access and audit trails fall under the Enterprise License
  • RAM and CPU requirements are not stated in the README
  • Documented quick start requires Docker
  • no GPU
  • Needs Docker
  • Models: OpenAI, Anthropic, Google, open-source models
  • port 5678
Documentation ↗ (opens in a new tab)Repository on GitHub ↗ (opens in a new tab)
282 out of 100

Activepieces

Self-hosted workflow automation with TypeScript integrations, alternative to Zapier

25.2k starscustom licenselast commit Oct 2026

Activepieces is a no-code workflow builder with loops, branches, auto retries, HTTP calls and npm-backed code steps, and flows are versioned. Integrations are called pieces: TypeScript npm packages, 280+ of which are exposed as MCP servers for Claude Desktop, Cursor or Windsurf. It also has AI pieces for several providers, human-in-the-loop approvals, and chat and form triggers.

Strengths

  • Pieces are open-source TypeScript npm packages with hot reloading for local development
  • 280+ pieces usable as MCP servers from Claude Desktop, Cursor or Windsurf
  • Flows are versioned and support loops, branches and auto retries
  • Built-in approval, delay, chat and form triggers for human-in-the-loop flows

Weaknesses

  • Enterprise features sit under a separate commercial license, not MIT
  • README does not state RAM, CPU or database requirements
  • Automation-first; AI agents are one feature rather than the core design
  • README claims of 200+ and 280+ pieces are inconsistent
  • no GPU
  • Docker + Compose
  • Compose runs PostgreSQL, Redis
  • Models: OpenAI
Documentation ↗ (opens in a new tab)Website ↗ (opens in a new tab)Repository on GitHub ↗ (opens in a new tab)
369 out of 100

banana-slides

AI slide deck generator that exports PPTX, PDF and narrated video

Live demo ↗ (opens in a new tab)15.9k starsAGPL-3.0last commit Oct 2026

Banana Slides generates presentations from a topic, outline, page descriptions or uploaded PDF/Docx/MD/Txt files, rendering each page as an image with image-generation models such as Nano Banana Pro. Pages can be edited by natural-language instructions or box selection, and exported as PPTX, PDF, editable PPTX (beta) or an MP4 with TTS narration. It runs as a Flask backend plus a web frontend and needs an API key for a text/image provider.

Strengths

  • Several providers: Gemini, OpenAI-format, Volcengine, Vertex AI and LazyLLM vendors like Qwen and DeepSeek
  • Exports PPTX, PDF, editable PPTX (beta) and narrated MP4 with subtitles
  • Edit pages by natural-language prompt or box selection; per-page templates supported
  • Prebuilt Docker Hub images with a docker-compose.prod.yml, plus a CLI and agent skills

Weaknesses

  • AGPL-3.0 license restricts use in closed-source hosted services
  • Needs paid external image-generation API; README warns Nano Banana Pro calls are costly
  • Editable PPTX export is beta; better results need a Baidu API key
  • README is mostly Chinese; desktop build is at release-candidate stage (v0.9.0-rc.7)
  • no GPU
  • Docker + Compose
  • Needs Gemini/OpenAI/Volcengine/Vertex or LazyLLM API key, FFmpeg with libass (video export), LibreOffice (optional, PPTX upload), Baidu API key (optional, editable export)
  • Models: Gemini (Nano Banana Pro), OpenAI-compatible (gpt-image-2), Volcengine Doubao/Seedream, Vertex AI, LazyLLM vendors (Qwen, DeepSeek, GLM, Kimi, MiniMax, SiliconFlow)
  • port 3011
Live demo ↗ (opens in a new tab)Documentation ↗ (opens in a new tab)Repository on GitHub ↗ (opens in a new tab)

Written from each project's README and checked facts. Spot something wrong? Report it on GitHub (opens in a new tab).