Documentation

Build diagrams
at the speed of thought

Everything you need to go from a plain-language description to a production-ready flowchart — in seconds. No installations, no account, no data stored.

3Steps to first diagram
0Data stored server-side
3Export formats
Diagrams you can create

Overview

AI Flowchart Studio is a browser-based diagram tool that converts plain-language descriptions into visual diagrams using Google's Gemini AI. You can also build diagrams from scratch with the Manual Builder — no API key required.

All diagrams are rendered using Mermaid.js — giving you clean, scalable output you can drop directly into GitHub, Notion, or any Mermaid-compatible tool.

AI-Powered Generation

Describe your process in plain English and get a rendered diagram in seconds

🔧

Manual Builder

Add nodes, connect them, label branches — full control without writing code

🖱️

Click-to-Edit Nodes

Click any rendered node to rename it or change its shape on the spot

📦

Multiple Export Formats

Mermaid code, SVG vector, or PNG raster — pick what fits your workflow

🔒

Private by Design

Your API key lives only in localStorage — never sent to our servers

💾

Project Saves

Save and reload multiple diagrams locally across browser sessions

Quick Start

Create your first AI-generated flowchart in under two minutes.

1

Get a free Gemini API key

Go to Google AI Studio at aistudio.google.com and create an API key. The free tier supports hundreds of generations per day — no credit card needed.

2

Paste your key into the app

Expand Gemini API Settings in the left panel, paste your key, and click Save Key. Stored in your browser only — do this once per device.

3

Describe your flow and generate

Type your process description in the AI Generate tab and click Generate Flowchart. Your diagram renders on the canvas in real time.

💡

No API key? Use Manual Mode

Switch to the Manual Builder tab to construct diagrams node-by-node without any API key.

API Key Setup

AI Flowchart Studio uses a BYOK (Bring Your Own Key) model. You supply your Gemini API key and the app calls Google's API directly from your browser.

Getting your key

1

Open Google AI Studio

Go to aistudio.google.com and sign in with your Google account.

2

Create an API key

Click Get API Key → Create API key. Select or create a Google Cloud project when prompted.

3

Copy and save in the app

Copy the key (starts with AIza…), expand Gemini API Settings, paste it, and click Save Key.

🔐

How your key is stored

Saved in localStorage in your browser. Only sent to Google's Gemini API when you click Generate — never passes through AI Flowchart Studio's servers.

⚠️

Keep your key safe

Don't share your API key or paste it into unknown sites. If compromised, revoke it in Google AI Studio and generate a new one.

AI Generation

The AI tab sends your prompt to Google Gemini, which returns valid Mermaid flowchart syntax. That code is rendered live on the canvas as an interactive SVG.

How it works

1

You write a prompt

Describe your process in the text area. Include start points, decision branches, and end states for best results.

2

Gemini converts it to Mermaid

Your prompt is sent to Gemini with a system instruction that constrains output to valid flowchart TD syntax only — no prose, no commentary.

3

Diagram renders on canvas

Mermaid.js parses and renders the result as a scalable SVG. Pan, zoom, click nodes to edit, and export — all immediately available.

After generation

  • Click any node on the canvas to open the node editor — rename it or change its shape.
  • The Mermaid Output field shows the raw code — edit it directly and re-render for precise tweaks.
  • Generate again with a refined prompt to replace the diagram entirely, or combine AI generation with manual edits.
💡

Generation failing?

Check that your API key is saved and valid. A network error in the status list may mean your free-tier quota is exhausted for the day.

Prompt Writing Guide

The quality of your diagram depends directly on the quality of your prompt. Gemini infers structure well — but the more explicit you are, the better the output.

What to include

  • Start node — Where does the process begin? (e.g. "Start when the user lands on checkout")
  • Actions — What happens at each step? Use verbs: validate, check, send, display, redirect.
  • Decision points — Where does the flow branch? Name both paths explicitly — the Yes/success path and the No/failure path.
  • End states — What are all possible termination points? Name each one.

Prompt examples

User authentication
Create a login flow. User enters email and password. Check if account exists — if not, show "account not found" and offer signup. If it exists, verify password — if wrong, increment failure count; after 3 failures lock the account. If correct and new device, send OTP — if OTP fails, reject. If OTP passes or device recognized, redirect to dashboard.
E-commerce order
Show an e-commerce order process. Start when customer clicks Place Order. Validate payment — if invalid, show error and retry. If valid, charge card — if declined, show failure. If charged, check inventory — if out of stock, refund and notify. If in stock, create order, send confirmation email, trigger dispatch, set tracking to Processing. End.
CI/CD pipeline
Document a CI/CD pipeline triggered on git push. Run unit tests — if failing, notify developer and stop. If passing, run integration tests — if failing, alert team and stop. Build Docker image, push to registry, deploy to staging. Run smoke tests — if failing, roll back and alert. If passing, await manual approval — if rejected, stop. If approved, deploy to production, run health check, post Slack notification.

Tips for better results

  • Use connecting words: if, then, otherwise, when, after, on success, on failure
  • Name each branch explicitly: "if valid → …" and "if invalid → …"
  • One action per node — don't bundle multiple steps together
  • State every end state so Gemini knows where to terminate each branch
  • For complex flows, break into sub-processes and generate separately

Iterate, don't perfect

Generate a first draft, see what comes out, refine the prompt. Two or three iterations produces a polished result faster than one long prompt upfront.

Manual Builder

Build diagrams node-by-node without AI. Useful for precise, predictable diagrams or when you don't have an API key.

Node types

Rectangle
Standard process step. Use for actions, tasks, or states. The most common type.
Diamond
Decision point. Use for Yes/No questions or branching conditions.
Circle
Terminal node. Use for the explicit start or end of a flow.
╭▭╮
Rounded Rectangle
Soft-edged step. Often used for start/end states in flowchart conventions.

Adding nodes

1

Switch to Manual Builder tab

Click Manual Builder in the left panel tab row.

2

Enter a label and choose a shape

Type the node text in Node Label, select a shape from the dropdown.

3

Click Add Node

The node appears in the list and renders on the canvas immediately.

Connecting nodes

  • Select the From node (origin) and To node (destination)
  • Optionally add an Edge Label — e.g. "Yes", "No", "On success"
  • Click Add Connection to draw the arrow
ℹ️

Mixing AI and manual

Generate a base with AI, then switch to Manual Builder to add or refine nodes. The Mermaid output field reflects all changes in real time.

Canvas Controls

The canvas is a scrollable, zoomable area that holds your rendered diagram.

Navigation

  • Pan — Click and drag the canvas background to move around.
  • Zoom — Use + / buttons in the bottom-right corner, or the zoom dropdown in the toolbar.
  • Fit to screen — Click Fit in the toolbar to reset zoom so the full diagram is visible.
  • Reset zoom — Click the zoom percentage to snap back to 100%.

Node interaction

  • Click a node to open the floating Node Editor — edit label or change shape.
  • Hover a node to see an indigo glow highlight.
  • Press Esc or click outside to dismiss the editor.

Toolbar reference

  • Layout — Switch flow direction: TD (top-down), LR (left-right), BT, RL.
  • Undo / Redo — Step backward or forward through edit history.
  • Clear — Wipe the canvas and start fresh.
  • Export — Open the export menu for Mermaid, SVG, or PNG.

Export & Share

Click Export in the canvas toolbar to open the export menu. Three formats available.

{ }

Mermaid Code

Copy to Clipboard

Raw Mermaid syntax. Paste anywhere that renders Mermaid natively.

GitHubNotionGitLabObsidian
⟨/⟩

Vector

Download SVG

Infinitely sharp at any size. Best for docs, design tools, and web embeds.

FigmaConfluencePrintWeb
🖼

Raster

Download PNG

High-resolution raster. Perfect for slides, Slack, and quick sharing.

SlidesSlackJiraEmail

Embed Mermaid in GitHub

GitHub natively renders Mermaid in Markdown. Paste your exported code inside a fenced block tagged mermaid:

Markdown
```mermaid
flowchart TD
    A([Start]) --> B[Enter Credentials]
    B --> C{Account exists?}
    C -->|Yes| D[Verify Password]
    C -->|No| E[Show Signup]
    D -->|Correct| F([Dashboard])
    D -->|Wrong| G[Show Error]
```
💡

Notion users

Create a /Code block, select Mermaid as the language, and paste. Notion renders it live.

Keyboard Shortcuts

ActionShortcutContext
Generate flowchartCtrl+EnterFrom prompt text area
UndoCtrl+ZReverts last change
RedoCtrl+Shift+ZRe-applies reverted change
Zoom inCtrl++Canvas focused
Zoom outCtrl+Canvas focused
Fit diagram to screenFCanvas focused
Close editor / overlayEscAny open popover
Save projectCtrl+SSaves to browser storage

On macOS, substitute Ctrl with .

FAQ

Is my API key sent to your servers?

No. Your key is stored only in your browser's localStorage. When you click Generate it goes directly from your browser to Google's Gemini API. AI Flowchart Studio's servers never see it.

Do I need to pay for the Gemini API?

The Gemini API has a generous free tier supporting hundreds of diagram generations per day — enough for personal and team use. Heavier usage may require a paid Google Cloud plan. Check Google AI Studio for current limits.

What diagram types can I create?

AI Flowchart Studio generates Mermaid flowchart diagrams — covering flowcharts, process maps, decision trees, system architecture diagrams, CI/CD pipelines, and user journeys. Sequence diagrams, Gantt charts, and ER diagrams are outside the current scope.

How are my diagrams saved?

Diagrams are saved as named projects in localStorage. They persist between sessions on the same browser and device but aren't synced to any cloud. Clearing browser data removes them — use the Mermaid code export as a portable backup.

The generated diagram looks wrong — what do I do?

Try refining your prompt with more explicit branching language. If the structure is mostly correct, click individual nodes to edit them. For large mis-renders, check the Mermaid output field for syntax errors, fix them inline, and re-render manually.

Can I use the Manual Builder offline?

The Manual Builder works without an internet connection once the page is loaded. AI generation requires connectivity to reach the Gemini API. Mermaid.js is loaded from a CDN, so the initial page load also needs connectivity.

Can I embed the exported SVG on my website?

Yes. The exported SVG is a standard vector file. Embed with an <img> tag, inline it in HTML, or import into Figma. Diagrams you create are yours — no attribution required.

Explore the app

AI Flowchart Studio is a browser-based tool. All diagram data stays on your device. Your Gemini API key is transmitted directly to Google — never to AI Flowchart Studio's servers.