Ultra-fast storyboards: minutes, not days
For producers and directors with a shoot to sell: ChatGPT turns the script into one column per frame on a Storyflow board, ready for your pictures.

Creatives working for these brands already got early access
Watch a real project come together, from a blank board to a finished plan, clicked through step by step.
Pick a board to see what you can build, then let AI fill it in. Every template is a real, editable starting point on the same infinite canvas.

Lay out your shots frame by frame with images and notes, then reorder them until the sequence reads like the finished cut.
Yes, ChatGPT can make a storyboard: it cuts a script into frames, writes the action, the voiceover and the seconds for each one, and can draw a picture for a frame in its own chat, but what comes back is a table and some loose pictures inside a conversation. Storyflow is an AI visual workspace that turns a script into a storyboard on one infinite canvas: a column for every frame, a picture space at the top of each, the action and the voiceover underneath, and arrows that keep the frames in order. Connected to Storyflow, ChatGPT builds those columns on your board itself, and because it runs on its own model the work spends none of your Storyflow AI credits.
ChatGPT is good at this even with nothing connected. Paste the script for a 30-second spot and it cuts it into frames, gives each one a shot size and a camera move, writes the voiceover that plays under it, and points out that your seconds add up to 34, not 30. Ask for pictures and it draws them, one frame at a time. The trouble starts when the storyboard has to leave the chat. The words sit in a table, the pictures are further up the conversation, and a face can change from one drawing to the next. To show a client, you paste it all into slides by hand, and again after every change.
Connect ChatGPT to Storyflow and ask for the storyboard on the shoot's board. If the script is a PDF, a Word file or a Final Draft file on that board, ChatGPT reads it from there. Then it builds the whole row in one request, up to 60 frames. Every frame is a column with the same five cards: an empty picture frame, a one-line subject with two to four tags for the shot size, the light and the camera, a "Shots required" checklist, a direction note for the action, the voiceover and the seconds, and a gear note. Arrows join the columns from left to right. Then the pictures go in, three ways: pictures already on the board are moved into their frames, pictures go in from direct image addresses ChatGPT has (a link to a web page is not a picture), or you send your own through an "Add your photos" link.
Here is a storyboard you can copy without an account. A 30-second spot for a bike shop, in eight frames. Frame 1, 3 seconds: the shop shutter rolls up at dawn. Frame 2, 4 seconds: a dusty bike is wheeled in. Frame 3, 4 seconds: hands straighten the wheel. Frame 4, 3 seconds: the chain drops onto the gear. Frame 5, 5 seconds: the rider sets off along the harbour, and the voice says "Spring starts with one ride." Frame 6, 4 seconds: she passes the café and a friend waves. Frame 7, 3 seconds: close on her grin. Frame 8, 4 seconds: the shop front, the logo and the web address. That is 3, 4, 4, 3, 5, 4, 3 and 4, which makes 30. The 15-second cut keeps frames 1, 5, 7 and 8: 3, 5, 3 and 4 seconds, which makes 15. On the board that is eight columns of five cards, so 40 cards. Three frames take the shop photos already on the board and the other five take the director's sketches through the photo link.
Every storyboard frame needs the same few things, whoever writes it. A number, so people can say "frame 5" on a call. A picture, even a rough one. One line for what we see, in the present tense. The words we hear under it, if any. Any words on screen. The shot size and the camera move. And how many seconds it lasts. Add the seconds up before anyone approves anything. A voice reads about two and a half words a second, so a 30-second spot holds about 75 words at most, and fewer once the last frame with the logo takes its share. Give that last frame its own time. If there is a shorter cut, mark which frames it keeps while the long one is still on the table.
Now the limits. Storyflow does not make pictures or search for them over the connector, so every frame starts empty. A picture ChatGPT draws in the chat, or one you attach there, does not reach Storyflow by itself: save it and send it through the "Add your photos" link, which works for 15 minutes and takes up to 200 pictures. ChatGPT cannot see the pictures on your board, only their web addresses and captions, so checking a drawing against its note is your job. It has no tool that deletes a card, and a storyboard you ask for twice lands beside the first one. Storyflow does not play the frames back as a timed film with sound. Storyflow's own built-in AI, inside the app, is the other route when you want the frames drawn for you.
Ultra fast from brief to board
The layout is the quick part. Most of your time goes on the pictures, where it belongs.
01
In ChatGPT on the web, open Settings, then Security and login, and switch on Developer mode. Open chatgpt.com/plugins and click the plus button. Name the connector Storyflow, choose Public endpoint, paste https://mcp.storyflow.so/mcp and create it. Sign in to Storyflow when it asks, pick the shoot's project and leave changes allowed.
02
Paste the script into the chat, or put the file on the shoot's board and ask ChatGPT to read it there. Say how long the film is and how many frames you want. For example: "A 30-second spot, eight frames. Build the storyboard on the Harbour Bikes board." A row of columns appears, one per frame, joined by arrows.
03
Each column shows an empty picture frame, the subject with its tags, the checklist of what to shoot, the direction note and the gear. Change a line by hand, or tell ChatGPT which frame to change and it edits that card where it sits.
04
Ask ChatGPT to use the photos already on the board, or to give you the "Add your photos" link. Pick your sketches and stills from your phone or computer and they fill the empty frames. When the client is ready to look, share the board from Storyflow with a view-only link.
ChatGPT writes the storyboard. Storyflow keeps it where the client and the crew can read it, frame by frame.

One column per frame
Every frame is its own column, with the picture space on top and the notes below it. Arrows join the columns in order, so a client reads the film from the first frame to the last.
See the storyboard template →
The pictures
Use the stills and location photos already on the board, add reference pictures from direct image addresses ChatGPT has (a page link will not do), or send your own sketches through the "Add your photos" link. A picture added from the web is copied into Storyflow, so it stays when the page it came from is gone.
See the AI storyboard generator →
The shoot
The "Shots required" checklist on a frame is the list of shots that frame needs on the day, and the gear note says what it takes. The crew ticks it as they go, so the storyboard the client approved is what the camera team works from.
See the shot list with Claude →
Any assistant
Storyflow's connector is one web address. Your director can add it to Claude while you work in ChatGPT, and both of you build on the same frames. Each person chooses which projects their own assistant may open.
See how to connect your assistant →ChatGPT runs on its own model, so a storyboard built over the connector uses no Storyflow AI credits. A free account (today, one invited by a paid member) gets a weekly allowance of actions from its assistant, and the pricing page shows the current number. Plus, Pro and Max have no limit. Storyflow is paid-only during early access, with plans from $7.99 a month billed annually ($9.99 monthly). A free plan launches before the end of 2026, and anyone a paid member invites to a board joins free.
No Storyflow AI credits spent on frames ChatGPT builds
No limit on connector actions on Plus, Pro and Max
The director and the client's team join the board free when you invite them
Script, frames, reference pictures and schedule on one board
You choose which shoots ChatGPT may open, and whether it may change them

Built for people with taste
ChatGPT is good at frames. The trouble is everything that happens after its answer.

The layout
Words and picture in one column: Each frame's picture sits on top of its own action, voiceover and seconds. Nobody has to work out which drawing goes with which line.
Up to 60 frames in one request: ChatGPT sends the whole row at once, so a two-minute film arrives as one finished row.
Start from a storyboard template: Ask ChatGPT to search Storyflow's templates for a storyboard and put one on the board. It arrives with its own example frames, and ChatGPT rewrites them for your script.
The script stays beside the frames: The PDF, Word or Final Draft file sits on the same board, and anyone can check a frame against the page it came from. ChatGPT reads the first 40,000 characters of a file, so a long script goes in parts.

The pictures
A link for your own frames: Ask for the "Add your photos" link, open it, and pick the sketches or stills from your phone or computer. It takes up to 200 pictures and works for 15 minutes.
Empty frames fill up as you upload: ChatGPT tells the link which empty frames to fill. Each picture goes into the next one as it arrives, so look over the order on the board when the upload is done.
Photos on the board become frames: Put the location stills on the board first and ask ChatGPT to use them. It moves each one into its column, so no empty box is left sitting beside it.
Reference pictures from the web: A picture goes in from a direct image address ChatGPT has, twelve per request (a link to a web page is not a picture), and Storyflow stores its own copy. Check you may use a picture before it goes to a client.

The sign-off
Change one frame by asking: The voiceover on frame 5 is now "Spring starts here." Tell ChatGPT, and it rewrites that one note. The other seven frames are not touched.
Two versions side by side: Ask for a second opening and the new frames land beside the first set. Show the client both, keep one, and remove the other by hand.
One link for the client: Share the board from Storyflow with a view-only link. The client opens it in a browser with no account and reads the frames in order. To leave a comment, they sign in.
Ask what changed: Ask ChatGPT what changed on the board this week. It reads the board's history and lists what was added, edited or moved, and by whom. What ChatGPT wrote for you shows under your name.
Every template opens as a real, editable board on the infinite canvas. Pick the closest fit and make it your own.
Lay out your shots frame by frame with images and notes, then reorder them until the sequence reads like the finished cut.

One app instead of five
Storyflow holds the frames ChatGPT writes. Over the connector it does not draw them for you, and it does not play them back as a film.
Recommended
Frames written from the script by AI
Frames laid out in order on a board the team shares
Script, reference pictures and frames on one open canvas
Shooting schedule built on the same board
Team edits a frame without asking the AI again
Reviewers comment on a single frame
AI pictures placed in each frame for you
Timed animatic with sound
Frames written from the script by AI
Frames laid out in order on a board the team shares
Script, reference pictures and frames on one open canvas
Shooting schedule built on the same board
Team edits a frame without asking the AI again
Reviewers comment on a single frame
AI pictures placed in each frame for you
Timed animatic with sound
Frames written from the script by AI
Frames laid out in order on a board the team shares
Script, reference pictures and frames on one open canvas
Shooting schedule built on the same board
Team edits a frame without asking the AI again
Reviewers comment on a single frame
AI pictures placed in each frame for you
Timed animatic with sound
Join early creators getting structured workspaces and AI that remembers their projects
“Storyflow has sped up my workflow by at least 3x, which means more flow state and more projects I can actually ship. It truly changed the way me and my team create.”

Reilin Joey
Director & YouTuber
“One prompt gets me a structured board. But the tactics are my favorite. I run my YouTube scripts through them and my intros and retention got better. It's amazing.”

Justkay
YouTuber & Freelance Filmmaker
“I used to juggle five apps to plan a project. Now I describe what I am making and get boards, lists, and a schedule. All in one place.”

George
@fernwehchronicles
What producers and directors ask before the storyboard goes to a client.
Yes. ChatGPT cuts a script into frames and writes the action, the voiceover and the seconds for each one, and it can draw pictures in its chat. Connected to Storyflow, it builds the storyboard on your board instead: a column per frame with a picture space, a subject line, a checklist of what to shoot, a direction note and the gear, joined by arrows. The pictures go in from your board, the web or your own files.
Storyflow gives ChatGPT a board to build the storyboard on: one column per frame with a picture space, the action, the voiceover, a checklist of what to shoot and the gear, in order, where the client and the crew read it. The same connector works with Claude, Gemini and Grok. Boords still plays the frames back as a timed animatic with sound, and its own AI draws a picture into each frame.
In its own chat, yes. On your Storyflow board, not directly: the connector does not make pictures, so each frame starts empty, and a drawing made in the chat does not travel to the board by itself. Save the drawings and send them through the "Add your photos" link ChatGPT gives you. Storyflow's own built-in AI, inside the app, is the other route if you want the frames drawn there.
Ask ChatGPT for the photo link for your Storyflow board. It replies with a link called "Add your photos". Open it on your phone or computer and pick the pictures. They fill the empty frames, and any extras land in a column of their own. The link works for 15 minutes and takes up to 200 pictures. It carries pictures only, not PDFs or video.
Use ChatGPT on the web. Under Settings, then Security and login, switch on Developer mode. Next open chatgpt.com/plugins, click the plus button, call the connector Storyflow, pick Public endpoint and paste https://mcp.storyflow.so/mcp. Create it and sign in to Storyflow. Storyflow's setup notes list ChatGPT Plus, Pro, Business, Enterprise and Edu, and on Business and Enterprise an admin switches Developer mode on.
Yes, when the file is on your Storyflow board. ChatGPT reads PDF, Word and Final Draft files from the board, up to 25 MB each, and gets the first 40,000 characters, so a long script goes in parts. A scanned PDF is not read. A file you attach in the ChatGPT chat is read there, but the file itself does not reach Storyflow.
No. Storyflow sends ChatGPT each picture's web address and its caption, not the picture. So ChatGPT can put a photo captioned "shop front" in the last frame, but it cannot tell whether a sketch matches its note or whether a face stays the same across frames. Give your pictures clear captions, and check the frames yourself before the client does.
Up to 60 frames in one request on a Storyflow board, each as its own column of five cards. For a longer film, ask for it in parts, one scene or one act at a time, and each part gets its own row. Pictures from web addresses go in twelve per request, and the photo link takes up to 200 of your own pictures.
Tell ChatGPT which frame to change and it edits that card in place on your Storyflow board. It has no tool that deletes a card, and it cannot replace a whole storyboard: ask for the storyboard again and the new one lands beside the old one. Remove the old one by hand. What you delete stays in the board's trash for 7 days.
Yes. The frames are ordinary cards on a Storyflow board, so they stay after the chat ends. Share the board from Storyflow yourself, because ChatGPT cannot create or change share links. A view-only link opens in a browser with no account. To comment, a person signs in. Anyone a paid member invites to the board joins free.
No. ChatGPT runs on its own model, so the storyboard it builds over the connector uses no Storyflow AI credits. A free Storyflow account (today, one invited by a paid member) gets a weekly allowance of actions from its assistant, and the pricing page shows the current number. Plus, Pro and Max have no limit.
Yes. Storyflow's connector is one address, https://mcp.storyflow.so/mcp, for every assistant. In Claude, add it under Customize, then Connectors; on the Free plan Claude allows one custom connector. Gemini's custom apps need a personal Google account with Keep Activity on, in the US, for people 18 or over, in English. On Grok Business and Enterprise a team admin adds it in the xAI console first. Each one builds the same frames.
Only the ones you choose. When you sign in, Storyflow asks whether ChatGPT may use all your projects or only the ones you tick, and whether it may make changes. Without changes it can only read, and it cannot give you the photo link. Every connection ends after 30 days, and then ChatGPT asks you to sign in again.
More from Storyflow
Storyflow MCP server
Connect ChatGPT to Storyflow
AI storyboard generator
Script to storyboard
Commercial storyboard
Shot list with Claude
Production plan with ChatGPT
Film treatment with ChatGPT
Call sheet with ChatGPT
Storyboard template
How to make a storyboard
Animatic maker
Storyboard examples
Video storyboard
Storyflow for producers
Storyflow for directors
One column per frame, your pictures on top, the words underneath, and one link for the client.