Is AI-Flow free to use?

Yes. The free tier includes 25 welcome credits and 20 free runs/day for BYOK nodes, inputs, and tools. No credit card required.

Where is my data stored?

Flow definitions are saved in the cloud. Generated media is kept for 7 days on the free tier and 30 days on higher tiers — download anything you need to keep.

Can AI-Flow generate images and video?

Yes. It supports image generation via Nano Banana 2, GPT Image 1.5, Seedream 4, Imagen 4, and Flux, plus video via Seedance, Kling, and Veo nodes.

How do I call a flow from my own application?

Use the API Builder to expose a flow as a REST endpoint. Define inputs, add webhooks for async completion, and retrieve outputs programmatically — the same flow can also be shared as a form-based app.

Can I self-host AI-Flow?

Yes, the project is open source on GitHub. Self-hosting means you manage the node server, your own provider keys, and storage — no platform fees, but no managed media retention either.

What are subflows and when should I use them?

Subflows are reusable workflow modules you embed inside larger flows. Use them to isolate repeated logic like product-photo preprocessing or prompt assembly, so you maintain one copy instead of duplicating nodes across canvases.

Directory/Automation/AI Workflow Builders/AI-Flow

AI-Flow

AI-Flow is a node-based canvas for chaining AI models — GPT, Claude, Gemini, Flux, Veo, Seedance, Kling and hundreds more — into a single repeatable workflow. Bring your own API keys and pay providers at cost, or run on platform credits with zero setup. Any flow can be exposed as a REST API, a webhook, or a stripped-down form app your teammates can run without ever seeing the canvas. Templates cover product mockups, batch image and video generation, storyboards, OCR and copy pipelines. Built for developers, indie founders and content ops teams who need multi-model automation without wiring the plumbing themselves.

Automation

Visit website

Added on

Nov 4, 2026

732 modelsBYOK keysVisual canvasREST APIWebhooksApp sharingTemplatesImage genVideo gen3D gen

Product information

Everything worth knowing about AI-Flow

The complete picture — from everyday features to the technical detail builders and IT folks dig for.

Core features

What it actually does

Multi-Model Workflows on One Canvas

Chain models from OpenAI, Anthropic, Google, Replicate and xAI inside a single visual flow to build end-to-end content pipelines instead of juggling five tabs and a spreadsheet.

Flow-to-App Publishing

Design a workflow once, choose which inputs are exposed, and hand it to your team as a simple form-based app via a share link. No canvas, no setup for the people running it.

Ready-Made Workflow Templates

Start from curated templates for product mockups, batch image generation, product video, storyboard-to-video, 3D asset generation and more — then adapt them to your own use case.

Bring Your Own API Keys (BYOK)

Plug in your own OpenAI, Anthropic, Replicate, Google or xAI keys and pay providers directly at their published rates. Or skip keys entirely and run on platform credits.

Image, Video and 3D Generation

Generate and edit images, video, speech, music and 3D assets from the same workflow — Flux, Seedream, Nano Banana, Veo, Kling, Seedance, Grok Imagine and others are wired in as nodes.

Technical capabilities

API Builder with Subflows and Webhooks

Expose any workflow as a REST endpoint, extract reusable subflows, and trigger flows from external systems via webhooks so the automation runs inside your own product.

Native Integration Nodes

Dedicated nodes for Notion, Airtable, Zapier and Make webhooks let you push generated output straight into the tools your team already lives in.

Must watch videos

How to Use Google Flow to Create Cinematic AI Videos (Step by Step)

11:42

Tutorial11:42

How to Use Google Flow to Create Cinematic AI Videos (Step by Step)

Watch Next 👉 https://www.youtube.com/watch?v=mKUVH610_3s&pp=ygUJcm9ib3ZlcnNl In this video, I show how to use Google Flow to build a cinematic AI video from consistent character and environment references, a six-shot storyboard, and individually animated scenes. The Google Flow Workflow: https://drive.google.com/file/d/1zDmWWCUhTA99KXkDoRK6HoT4FVurHum4/view?usp=sharing for inquiries: roboverse at youripartnerships.com #Roboverse

Create with Flow | How to use Google’s AI Creative Studio

3:10

Tutorial3:10

Create with Flow | How to use Google’s AI Creative Studio

Built with and for creatives, Flow is your AI creative studio powered by Google DeepMind’s most advanced models: Veo, Nano Banana, and Gemini. In this tutorial, you will learn to seamlessly create, refine, and compose your stories. With Flow, you can generate and edit images and videos in a single workflow, allowing you to build cohesive narratives with consistency. #Flow #AIFilmmaking #GenerativeAI We are so excited to see what you create. Create with Flow at https://flow.google/ See our Flow playlist on YouTube: https://www.youtube.com/playlist?list=PL590L5WQmH8cZy80Ko7dzpH2MyvNJFE-p #Flow #AIFilmmaking #GenerativeAI #GoogleDeepMind Subscribe to our Channel: https://www.youtube.com/google Find us on X: https://twitter.com/google Watch us on TikTok: https://tiktok.com/@google Follow us on Instagram: https://www.instagram.com/google Join us on Facebook: https://www.facebook.com/Google

Google Flow AI Filmmaking: Everything You Need to Know

20:53

Tutorial20:53

Google Flow AI Filmmaking: Everything You Need to Know

How to use Google Flow. Google Flow is Google's AI filmmaking tool, and the agent inside it will storyboard your entire project with you before you spend a single credit on video. In this video, you'll learn how to plan a video with AI, generate images and clips, keep a character consistent across every scene, and edit it all into a finished piece you can export. We go from a blank prompt to a finished 30 second ad, and I show you what each step costs along the way. You'll see how credits actually work, where the settings live that decide what you spend, and the one agent instruction trick that stops the AI from putting wrong text on your product. We also cover image editing with Nano Banana 2, frame to video, pulling new camera angles out of a clip you already have, building a reusable character, and fixing one small detail in a clip without regenerating the whole thing. If you're brand new to AI video, this is the walkthrough to start with. Timestamps: 0:00 Intro 0:29 Flow homepage tour 0:54 How credits work (and school or work accounts) 1:38 The workspace layout 2:15 Settings: image and video defaults 3:13 Agent instructions (brand rules that stick) 3:47 Your first prompt and the agent's questions 4:52 Adding a character and building the storyboard 6:56 Generating and editing images in Flow 8:07 Why AI still gets small text wrong 9:52 Generating clips and the cost approval step 11:18 Reviewing clips and catching continuity slips 12:04 Frame to video: start and end frames 13:25 New angles, new people, new styles from one clip 14:55 Characters and avatars for consistency 16:19 Fixing one detail without regenerating the clip 17:04 Editing your ad in Scenes 18:12 Exporting, downloading, and sharing 19:02 Tools and Storyboard Studio 20:31 Final thoughts Google Flow: https://flow.google Subscribe: https://www.youtube.com/@TeachersTech?sub_confirmation=1 Instagram: https://www.instagram.com/teacherstechlab LinkedIn: https://www.linkedin.com/in/teacherstech

Who uses it

Real use cases, no hype

Product Photo Generation

Batch-generate branded product shots and mockups from a single photo input using ready-made templates.

Multi-Model Content Pipelines

Chain Claude for copy, Flux for visuals, and Veo for video in one automated flow.

Image & Video Generation

Run text-to-image, image-to-video, and upscaling nodes from one canvas with any provider.

Internal Team Apps

Wrap a workflow in a form and hand it to non-technical colleagues via a simple link.

API Automation

Expose flows as REST endpoints and trigger them from your own product or backend.

Content Ops & Webhooks

Pipe generated assets into Notion, Airtable, Zapier, or Make automatically.

Storyboarding to Video

Turn script beats into cinematic video clips with consistent characters using Seedance/Kling.

What's great

Free tier, no subscription — pay providers directly with BYOK
700+ models in one canvas: GPT, Claude, Gemini, Flux, Kling, Veo
Drag-and-drop node editor, zero code to start
Turn any flow into a shareable form-based mini app

Technical strengths

REST API builder plus webhooks on every workflow
Subflows let you modularize and reuse pipeline logic
Open source on GitHub — self-host if you don't trust the cloud

Where it falls short

Not a coding agent — no codebase-aware generation
Generated media wiped after 7 days on free tier
Platform credits burn fast on video models
Canvas becomes spaghetti on 20+ node workflows

Technical limitations

No native git-style version control or flow diffing
BYOK buckets capped at 20 free runs/day
Weak retry/error-handling primitives on failed nodes

For technical folks

The deep-dive specs

Architecture

Node-based visual workflow canvas, cloud-hosted web application

Model library

732 models available

Supported providers

OpenAI, Anthropic, Google, Replicate, xAI

API access

REST API endpoints and webhooks per workflow, plus reusable subflows

Integrations

Notion, Airtable, Zapier, Make

Media retention

Generated media stored 7 days on free tier, 30 days on higher tiers

Source code

GitHub repository (DahnM20/ai-flow)

FAQ

Includes technical Q&A

Yes. You can plug in your own provider keys and pay them directly at their published rates. Each day you get a set number of free runs; beyond that you pay a small platform fee for storage and compute.