# Zephly > Your AI content team. In one canvas. Turn one brief into article + image + video + voiceover, in one reproducible workflow. Free tier, EU-hosted, MCP-native, 31 AI plugins on a visual DAG canvas. Zephly is a visual AI workflow automation platform. Users design directed acyclic graphs (DAGs) of AI nodes — text generation, image generation, video generation, audio synthesis, logic, and publishing — connect them via typed ports, and execute the workflow to produce reproducible, multi-format content. The product is aimed at solo creators, marketing teams, agencies, and small businesses who need to ship content in multiple formats (article + cover image + short video + voiceover) from a single brief, without writing code or stitching together five SaaS tools. Pricing is credit-based and transparent: 1 credit = 1 AI action. Logic and input nodes are free. AI nodes consume between 1 and 40 credits per call depending on the model (text is cheap, video is expensive). Plans are Free ($0 with 500 welcome credits + 100 per month), Starter ($15/mo, 500 credits), Pro ($39/mo, 2,000 credits), and Business ($89/mo, 5,000 credits). Annual billing saves 20% on every paid plan. There is no per-seat fee on the standard plans, no surprise overage billing — when you run out of credits the workflow simply waits until you top up or your monthly grant arrives. The platform is hosted in the EU (GDPR-compliant), supports the Model Context Protocol (MCP) natively for IDE and agent integration, and exposes a public REST API for programmatic workflow control. The frontend is Next.js 15 / React 19 / Shadcn UI; the backend is FastAPI / SQLAlchemy / PostgreSQL (Neon); storage is Cloudflare R2. Zephly differentiates from generic automation tools (Zapier, n8n, Make) by being AI-content-native — every plugin is an AI model or a logic block built around AI generation, with first-class typed ports for text/image/video/audio. It differentiates from developer frameworks (LangChain, LangGraph) by being a no-code visual canvas with a managed runtime. It differentiates from image-only generators (ComfyUI) by covering the full content stack (text + image + video + voiceover) in a single workflow. ## Product - [Home](https://zephly.ai/): Landing page. Wedge "Your AI content team. In one canvas.", animated workflow demo, plugin grid (31 plugins), pricing teaser (4 plans USD), and FAQ. Primary call to action: start a free workflow. - [Product overview](https://zephly.ai/product): Product page explaining the canvas, the typed ports system, the plugin catalog, and reproducibility. Includes screenshots of the editor and example output artifacts. - [Workflow editor](https://zephly.ai/product/workflow-editor): Deep-dive into the visual DAG editor. Nodes are dragged onto a canvas, ports are typed (text, image, video, audio, flexible), edges enforce type compatibility. Each node has a config schema, an execution status, and produces artifacts visible in the side panel. - [Plugins](https://zephly.ai/product/plugins): The 31-plugin catalog. Categories: Input (TextInput, ImageInput, VideoInput, AudioInput), AI Text (Claude Sonnet/Haiku/Opus, Gemini 2.5 Flash/Pro), AI Image (NanoBanana, Imagen 4), AI Video (Veo 3.1 Fast/Pro, Sora 2/Pro), AI Audio (ElevenLabs Flash/Multilingual/V3), Logic (Switch, Merge, Loop), Publishing (HTTP, S3, webhook). - [Integrations](https://zephly.ai/product/integrations): MCP server (20 tools to control workflows from any MCP client like Claude Code, Cursor, Cline), public REST API (workflows / runs / artifacts / credits / templates), webhooks for run completion, scheduled runs via cron expressions. - [Pricing](https://zephly.ai/pricing): Full pricing page with the 4 plans, monthly/annual toggle (-20% annual), feature matrix, FAQ on credits and refunds. JSON-LD SoftwareApplication schema with all four offers. - [Changelog](https://zephly.ai/product/changelog): Release notes — new plugins added, model versions upgraded, editor improvements, infrastructure changes. ## Templates - [Templates gallery](https://zephly.ai/product/templates): Public templates catalog. Each template ships with a name, description, category, the full graph (nodes + edges + viewport), and example_results — actual artifacts (text and image) from a successful run, so every node is clickable with a preview before you copy the template into your workspace. Users can copy any public template into their own workflows in one click. ## Solutions - [Solutions overview](https://zephly.ai/solutions): Top-level page routing visitors to the audience-specific landing page that matches them. - [Creators](https://zephly.ai/solutions/creators): For solo creators (YouTubers, newsletter authors, indie writers) who need to produce multi-format content from one brief — article + cover image + short video + voiceover — without juggling five tools. - [Marketing teams](https://zephly.ai/solutions/marketing-teams): For in-house marketing teams running repeatable content pipelines, brand-consistent assets at scale, and weekly editorial rituals. Templates and credit pooling are key. - [Agencies](https://zephly.ai/solutions/agencies): For agencies producing client deliverables. Workflows can be cloned per client, branded, and re-run on a schedule. Business plan covers most agency volumes; for higher volumes contact sales. - [Enterprise](https://zephly.ai/solutions/enterprise): For enterprises needing SSO, audit logs, dedicated SLAs, custom data residency, or private cloud. Pricing is custom — contact via /company/contact. - [Use cases](https://zephly.ai/solutions/use-cases): Concrete examples — LinkedIn carousel from a topic, product demo video from a product description, blog post + cover from a research brief, voice ad from a script. ## Compare - [vs n8n](https://zephly.ai/compare/vs-n8n): n8n is a general-purpose automation platform with optional AI nodes — strong for SaaS integration, weaker for AI content production. Zephly is purpose-built for AI content with typed AI ports, managed model access, and a credit billing model that maps to AI calls. - [vs Zapier](https://zephly.ai/compare/vs-zapier): Zapier connects 6,000+ SaaS triggers/actions and is excellent for cross-app automation; AI is a recent add-on. Zephly is AI-native — every node is an AI model or a logic block built around AI generation. Different category, different use case. - [vs Make](https://zephly.ai/compare/vs-make): Make (formerly Integromat) is a broader integration platform with visual scenarios. Zephly is narrower and deeper on AI content — better typed ports, better model selection, better artifact preview. - [vs LangGraph](https://zephly.ai/compare/vs-langgraph): LangGraph is a Python framework for building agent state machines — code-first, developer-targeted. Zephly is no-code, visual, managed — content team-targeted. - [vs LangChain](https://zephly.ai/compare/vs-langchain): LangChain is a developer toolkit for chaining LLM calls in code. Zephly is the no-code product layer above — for users who want results, not a framework. - [vs ComfyUI](https://zephly.ai/compare/vs-comfyui): ComfyUI is a powerful node graph for Stable Diffusion image generation, self-hosted. Zephly covers text + image + video + audio in one managed canvas, with hosted state-of-the-art models (Veo 3, Sora 2, Claude, Gemini). - [vs Flowise](https://zephly.ai/compare/vs-flowise): Flowise is an open-source self-hosted LLM-app builder, more dev-leaning. Zephly is a managed product with billing, hosting, and model access included. ## Documentation - [Docs home](https://zephly.ai/docs): Documentation index — getting started, editor reference, plugin catalog, templates, billing, API. - [Getting started](https://zephly.ai/docs/getting-started): Sign up, create your first workflow, run it, see artifacts. Includes MCP setup for Claude Code / Cursor / Cline. - [Workflow editor reference](https://zephly.ai/docs/workflow-editor): How nodes, ports (text/image/video/audio/flexible), and edges work. Validation rules, viewport, layout conventions. - [Plugins reference](https://zephly.ai/docs/plugins): All 31 plugins with their config schemas, ports (input/output direction), model lists, credit cost. Includes per-plugin examples. - [Templates](https://zephly.ai/docs/templates): How to copy a public template, customize it, and publish your own (with example_results so every node is previewable). - [Billing](https://zephly.ai/docs/billing): Credit system, plan grants, top-ups, refund rules, invoice access, payment methods. - [API](https://zephly.ai/docs/api): REST API reference. Endpoints for workflows, runs, artifacts, credits, templates. Authentication via API token. Same surface as the MCP server. ## Blog - [Blog](https://zephly.ai/blog): Tutorials, build logs, model comparisons, AI content workflow tips. Updated regularly. - [Blog feed](https://zephly.ai/blog/feed.xml): RSS/Atom feed for the blog, suitable for feed readers and AI assistants that ingest RSS. ## Trust - [About](https://zephly.ai/company/about): Company mission, team, founding story, why we build Zephly. - [Contact](https://zephly.ai/company/contact): Sales, support, partnerships, press. Email and contact form. - [Terms of Service](https://zephly.ai/legal/terms): Service terms, acceptable use, liability. - [Privacy Policy](https://zephly.ai/legal/privacy): Data handling, EU hosting (GDPR-compliant), data retention, sub-processors, user rights. - [Refund Policy](https://zephly.ai/legal/refund): Credit refund rules, subscription cancellation, prorations. ## Optional - [Brand](https://zephly.ai/company/brand): Logos, color palette (teal/cyan accents on dark backgrounds), typography, usage guidelines, downloadable assets. - [Partners](https://zephly.ai/company/partners): Integration partners, ecosystem allies, agencies and consultancies certified on Zephly.