Documentation

Everything GOGUL can do — how each tool works, what it costs, and which plan unlocks it. Prices and limits on this page are read directly from the product configuration.

Welcome to GOGUL

GOGUL is an AI platform for images, video and storyboards. You describe the shot, GOGUL orchestrates the right model — and keeps your characters consistent from frame to frame.

Every generation is paid in credits. You start with 100 credits on the Free plan, and the exact price of any action is always shown before you confirm it.

Quick start guide

Four moves from a blank page to a finished clip:

1

Generate an image

Open Image in the sidebar, type a prompt, pick a model and press Generate. Free-plan models start at 3 credits per image.

2

Save a character

On any image card choose Save as Character, or upload a photo on the Characters page. Generate Character Views so the AI sees your character from several angles.

3

Build a storyboard from a script

Paste a script on the Script page — analysis is free. Review the scene breakdown, attach your characters, then generate frames at 10 credits each on GOGUL 1.0 or 20 on GOGUL 3.0 with Character Lock.

4

Animate a frame

Open Video → Image to Video, pick a frame from your library and describe the motion. The Generate button shows the exact credit price before you commit.

GOGUL dashboard showing the list of generated storyboards
GOGUL dashboard showing the list of generated storyboards

Klappi assistant

Klappi is the built-in help assistant. It answers questions about GOGUL from this documentation and takes your account into account: your plan, your credit balance and the page you are on.

  • Where to find it: The round button at the bottom right of every dashboard page. On phones it is hidden on the Image page, where it would cover the prompt bar.
  • Daily limit: 20 questions per day per account. The window is rolling, not calendar-based: when the limit is reached, the panel shows how long until the next question.
  • What it does not do: Refunds, payments, chargebacks and account blocks — Klappi does not resolve them. These go to the support team.
  • If the answer did not help: Press “Write to support” right in the panel. The message goes to the support team, and the reply comes by email.

Image generation

The Image page (sidebar → Create → Image) turns a text prompt into pictures. Pick a model, a quality tier, an aspect ratio and a style; optionally attach reference images. The counter next to the Generate button sets how many images to make, and the button shows the total price before you press it.

Models and quality tiers locked below your plan unlock after any credit-pack purchase. A pack also lifts the storyboard frame limit to 12 and the character limit to 20 when your plan's limits are lower.

Image models

Every model, its price per image at 1K, the plan that unlocks it, and whether it accepts reference images.

ModelCreditsPlanReference images2K / 4K
FLUX.2 [pro]Exceptional photorealism & artistic images10All plans—
Seedream 5.0NewLatest generation & consistency10All plansup to 10
Seedream 4.5BudgetGreat value — unified generation & editing10All plansup to 10
Krea V2 MediumBudgetKrea · Fast & affordable10All plans—
Nano Banana LiteFastFast & cheap — Google Gemini 3.1 Flash Lite10All plansup to 5
FIBO 1.5NEWPrecise & predictable — strong photorealism, textures & in-image text12All plans—
FLUX 3NewPrecise layout & in-image text, up to 10 references14All plansup to 10
FLUX 3 KleinNewFast & cheap — layout, in-image text, up to 10 references3All plansup to 10
Krea 3NewKrea · Quality variant — reasoning, up to 8 references7All plansup to 8
Krea 3 FastFastKrea · Speed variant — same price, up to 8 references7All plansup to 8
Seedream 5.0 ProNewFlagship detail & realism15Starter and upup to 10
Krea V2 LargeNewKrea · Best quality & composition20Starter and up—
GPT Image 1.5Best for textPerfect text rendering & instruction following10Starter and upup to 4
Nano Banana 2.1RecommendedKeeps identity from reference — Google Gemini20Starter and upup to 14
GPT Image 2NewBest text rendering & photorealism — OpenAI20 on Starter · 30 on Pro · 30 on Studio · 30 on AgencyStarter and upup to 4
GPT Image 2.6 FlareNewOpenAI · Fast drafts & everyday images, up to 16 references13Starter and upup to 16
GPT Image 2.6 SunburstNewOpenAI · Detailed work that holds across edits, up to 16 references13Starter and upup to 16
Nano Banana ProPremiumBest photorealism — Google Gemini 3 Pro40Pro and upup to 14

Quality tiers

Higher output resolution adds a surcharge on top of the model price and requires a plan. Only models marked in the 2K / 4K column actually deliver larger pixels — the tier is hidden for the others.

QualityExtra creditsPlan
1K—All plans
2K+10Starter and up
4K+20Pro and up

Seedream 5.0 Pro is priced per quality tier instead of base + surcharge: 1K — 15, 2K — 30. Each reference image beyond the first adds one credit.

Styles

A style appends a proven prompt suffix so you don't have to write lighting and lens language yourself. The Free plan can use 3 styles (None, Cinematic, Photorealistic); the other 7 show a lock and open the upgrade dialog. This gate is plan-only — a credit pack does not unlock styles.

NoneCinematicPhotorealisticAnime3D RenderCyberpunkFantasyWatercolorOil PaintingComic Book
Several images per request
The ± counter next to Generate requests up to 4 variations of the same prompt at once, on every plan. Each image is billed at the model price plus any quality surcharge, so the button shows count × price. If some images fail, only those are refunded.
Enhance prompt
The sparkle button sends your prompt (and, in the editor, the image) to a language model that rewrites it into a detailed English description — scene, lighting, camera angle, atmosphere, palette, style. The result replaces your text, so edit it if it went too far. Free, never charged; rate-limited like generations.
Reference images
Attach a photo to keep a face, product or style. The slot count depends on the model — see the table. Models marked “—” (FLUX.2 [pro], Krea V2 Medium, FIBO 1.5, Krea V2 Large) generate from text only: their + button is disabled with a hint, the reference block is hidden, and switching to one of them drops the references you had attached. Switching to a model with fewer slots trims the list to fit.
After generation
Every result lands in your gallery. From a card you can download the image, upscale it, save it as a character, publish it to Community, or open it in the editor.

Image editor

The Edit page (sidebar → Tools → Edit) modifies an existing image instead of creating a new one. Upload a file or pick anything from your library; the tool rail appears once an image is loaded. Every tool works on the current result, so edits can be chained.

Text edits run on Seedream 5.0 Pro — 10 credits at 1K, 20 at 2K.

Tools

ToolWhat it doesCredits
Edit by textDescribe the change in the prompt bar — the model modifies only what you ask and keeps the composition. Model and quality are chosen in the same bar.10
Quick presetsChips that fill the prompt with a ready instruction. They do not run by themselves — press Generate, and the edit is billed as a normal text edit.10
Brush (inpaint)Paint a mask over the area and describe what should be there; brush size, undo and clear are in the panel.15
Select areaDrag a rectangle and choose whether to edit only that area or keep it untouched; the selection is added to your prompt. Available on Seedream 5.0 Pro only.10
Remove backgroundOne click — cuts the subject out on a transparent background.5
OutpaintExpand the canvas to a new aspect ratio; the model fills in the missing edges.12
UpscaleEnlarge to 2K or 4K — same prices as the Upscale section.10 / 15
RelightPick a light direction and one of the light presets (golden hour, blue hour, studio, dramatic, neon, soft daylight) to relight the scene.12
EnhanceOne click — sharpens and adds detail while keeping the content identical. Always runs on Seedream 5.0 Pro at 1K, whatever model is selected.10
Move (pan & zoom)Zoom into the image and drag to pan while you inspect it. It does not move objects and costs nothing.Free
Quick presets
Change background, enhance quality, style transfer, remove objects, change mood, change outfit, change scene. Each chip writes a prompt; you can edit it before generating. There is no separate face-swap tool — to replace a face, switch the edit model to Nano Banana 2.1 and describe the swap.

Outpaint works on images that are already in your gallery. For a fresh upload, run any other edit first (the result is saved) and then outpaint it.

Style reference while editing
Some models accept a second image as a style reference during an edit: Seedream 5.0 Pro, Seedream 5.0, Seedream 4.5, Nano Banana 2.1, GPT Image 2, Nano Banana Pro. For the others the Style Reference block is hidden.

Upscale

Upscale sharpens and enlarges any generated image or storyboard frame. Available from the image card, the editor and the storyboard editor.

  • 2K — 10 credits
  • 4K — 15 credits

The upscaled version is added to your gallery next to the original.

Video Generation Guide

Modes

The Video page generates short cinematic clips from a text prompt, a single image, a pair of start/end frames, or a set of reference images. This guide covers every mode, every model, and exactly how pricing works — so you always know what a clip will cost before you press Generate.

Video generation is available on every plan — advanced models unlock with a plan or a credit pack
Vidu Q4 is available to everyone, including the Free plan. The other models unlock at their plan tier (see the table below) — or with any credit pack purchase, which unlocks every model except Veo 3.1 Fast. If a model appears locked in the model picker, that's the reason — see the FAQ below.

Quick Start

From idea to finished clip in five steps:

1

Open the Video page

Go to Video in the sidebar. Your last unfinished setup is restored automatically — drafts are autosaved.

2

Pick a mode tab

Choose Text to Video, Image to Video, First & Last Frame, or Reference, depending on what you're starting from.

3

Write your prompt

Describe the shot — up to 7,000 characters. Focus on motion: what the camera does and what happens in the frame.

4

Choose model, duration, and resolution

The exact credit price updates live on the Generate button as you change settings — no surprises.

5

Hit Generate

Credits are charged when generation starts. Clips on Wan 2.7, Kling 3.0, Veo 3.1 Fast are usually ready within a few minutes; Vidu Q4, Wan 3.0, Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5, FLUX 3 Max, H3 Max · 3D Camera run through a queue and can take up to 25 minutes (Vidu Q4 — 25; Seedance 2.5 — 20; Wan 3.0, FLUX 3 Max — 15; Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, H3 Max · 3D Camera — 10).

The Four Modes

Each tab on the Video page is a different way of telling the model what to make. Pick the one that matches your starting material.

Text to Video
The simplest mode: write a prompt and get a video. Quick Presets — Cinematic, Anime, Realistic, Noir, and Watercolor — append a style description to your prompt with one click, so you don't have to write style language yourself.
Image to Video
Upload an image as the Start Frame and describe the motion in the prompt. You can also pick an image from My Images / Library — any of your previous generations — which is exactly how a storyboard frame becomes a video shot.
First & Last Frame
Provide two frames and a prompt describing the transition between them; the model animates from the first frame to the last. Supported by Wan 2.7, Kling 3.0, Veo 3.1 Fast.
Reference
Attach reference images and point to them in the prompt with tags like @Image1 — the model keeps the subject, location or style from the references while composing a new scene. Runs on Vidu Q4, Wan 3.0, Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5.
Motion Transfer · Separate tab
A photo of your character plus a short video of someone moving — the Dreamactor M2 model transfers the motion onto your character. Priced per second of source video; the exact cost is shown on the Generate button. It has its own tab next to the four modes above.
GOGUL Video Generator page with model picker, prompt field, and generated clips
GOGUL Video Generator page with model picker, prompt field, and generated clips

Models

Each model has its own plan requirement, supported inputs and price per second. Use this table to pick before you prompt.

ModelPlanWith audio · cr/sec at 720pAudio offInputsUnlock
Vidu Q4All plans25—Image, ReferenceAll plans
Wan 2.7Starter and up18—Text, Image, First & last frameUnlocks with any credit pack
Wan 3.0Starter and up18—Text, Image, ReferenceUnlocks with any credit pack
Seedance 2.0 ProStarter and up50—Text, Image, ReferenceUnlocks with any credit pack
Seedance 2.0 FastStarter and up15—Text, Image, ReferenceUnlocks with any credit pack
Seedance 2.0 MiniStarter and up10—Text, Image, ReferenceUnlocks with any credit pack
Seedance 2.5Starter and up58—Text, Image, ReferenceUnlocks with any credit pack
FLUX 3 MaxStarter and up43—Text, Image, KeyframesUnlocks with any credit pack
H3 Max · 3D CameraStarter and up20—3D CameraUnlocks with any credit pack
Kling 3.0Pro and up3024Text, Image, First & last frameUnlocks with any credit pack
Veo 3.1 FastStudio and up3831Text, Image, First & last framePlan-exclusive — no credit-pack bypass

The audio toggle changes the price on Kling 3.0, Veo 3.1 Fast: audio off is cheaper. The other models charge the same either way.

Models with Reference

6 models can also run the Reference tab: Vidu Q4, Wan 3.0, Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5. They differ in price per second, resolutions, durations and plan:

Tiercr/sec at 720pResolutionsDurations, sPlan
Vidu Q4NewUp to 12 references, 16s and 4K, by Shengshu25540p / 720p / 1080p / 2K / 4K3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16All plans
Wan 3.0NewUp to 30s clips with native audio, by Alibaba18480p / 720p / 1080p4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 30Starter and up
Seedance 2.0 ProProTop-tier quality up to 1080p, by ByteDance50480p / 720p / 1080p4, 5, 6, 8, 10, 12, 15Starter and up
Seedance 2.0 FastFastQuick turnaround generation, by ByteDance15480p / 720p4, 5, 6, 8, 10, 12, 15Starter and up
Seedance 2.0 MiniNewFast, budget-friendly generation, by ByteDance10480p / 720p4, 5, 6, 8, 10, 12, 15Starter and up
Seedance 2.5NewFlagship quality up to 1080p and 30s, by ByteDance58480p / 720p / 1080p4, 5, 6, 7, 8, 9, 10, 12, 15, 20, 25, 30Starter and up
  • 1080p is offered on Vidu Q4, Wan 3.0, Seedance 2.0 Pro, Seedance 2.5 only; the other tiers stop at 720p and a stale 1080p choice snaps back to 720p when you switch to them.
  • Wan 3.0, Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5 unlock on Starter — or with any credit pack purchase on lower plans.
  • On the Free plan Vidu Q4 renders 540p clips of up to 5 s — 60 credits for the longest one.
  • Vidu Q4 by plan: 720p, 1080p, 2K — Starter; 4K — Pro; clips longer than 5 s — Starter. Any credit pack purchase lifts these limits.
  • The Generate Audio toggle does not change the price of these tiers.

Pricing

Every video price comes from one formula: credits = base rate per second × seconds × resolution multiplier. Resolution multipliers are ×0.5 for 480p, ×1 for 720p, and ×2.25 for 1080p. “Auto” duration is billed as 5 seconds.

Resolution multipliers

  • 480p — ×0.5
  • 720p — ×1
  • 1080p — ×2.25
  • “Auto” duration is billed as 5 seconds.
  • Exception: Vidu Q4, Seedance 2.5, FLUX 3 Max, H3 Max · 3D Camera do not use these multipliers — each resolution has its own price, in credits per second: Vidu Q4: 540p 12, 720p 25, 1080p 31, 2K 49, 4K 100; Seedance 2.5: 480p 25, 720p 58, 1080p 141; FLUX 3 Max: 720p 43, 1080p 73; H3 Max · 3D Camera: 480p 13, 768p 20, 1080p 40.

Base rates (credits per second at 720p)

  • Vidu Q4 — 25
  • Wan 2.7 — 18
  • Wan 3.0 — 18
  • Seedance 2.0 Pro — 50
  • Seedance 2.0 Fast — 15
  • Seedance 2.0 Mini — 10
  • Seedance 2.5 — 58
  • FLUX 3 Max — 43
  • H3 Max · 3D Camera — 20
  • Kling 3.0 — 30 (24 audio off)
  • Veo 3.1 Fast — 38 (31 audio off)

What typical clips cost

SetupCredits
Vidu Q4 — 5s, 720p125
Wan 2.7 — 5s, 720p90
Wan 3.0 — 5s, 720p90
Seedance 2.0 Pro — 5s, 720p250
Seedance 2.0 Fast — 5s, 720p75
Seedance 2.0 Mini — 5s, 720p50
Seedance 2.5 — 5s, 720p290
FLUX 3 Max — 5s, 720p215
H3 Max · 3D Camera — 5s, 768p114
Kling 3.0 — 5s, 720p, audio on150
Veo 3.1 Fast — 5s, 1080p, audio on428
You always see the price first
The exact credit cost is shown on the Generate button and updates live as you change model, duration, resolution, or audio — before anything is charged.

Advanced options

Below the prompt you'll find extra controls. They're useful shortcuts, but it helps to know what they actually do.

Motion Intensity (0–10)
How much movement you want in the shot, from near-static to highly dynamic.
Camera Movement
Static, Pan, Zoom, Dolly, or Orbit — picks a camera move for the shot.
Generate Audio
Sits on the settings row next to duration and resolution, not inside this panel. Toggles sound on or off. On Kling 3.0, Veo 3.1 Fast, turning audio off also changes the price — see the models table.
Aspect Ratio
16:9, 9:16, 1:1, or 4:3. Note: Wan 2.7 ignores the aspect ratio setting.
Motion and Camera are prompt sugar
Motion Intensity and Camera Movement are appended to your prompt as plain text instructions — they don't flip any hidden model switches. Writing the same thing in your own words in the prompt achieves the identical result.

Reference mode

Reference (a tab on the Video page) composes a new scene from reference images (up to 12 — Vidu Q4; 9 — Wan 3.0, Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5) while keeping the subject, location or style they show. Attach images from a file, your library, a saved character, a location or a style, then point at them in the prompt.

Tags are positional: @Image1 is the first chip, @Image2 the second, and so on. Type @ in the prompt to pick a reference from a menu; focusing an empty prompt with references attached pre-fills the tags for you.

  • Runs on Vidu Q4, Wan 3.0, Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5. If another model is selected when you open the tab, the page silently switches to Wan 3.0 (the cheapest); leaving the tab keeps that model.
  • Up to 5 reference audio tracks can be attached, up to 15 MB each: MP3, 3–12 s — Vidu Q4; MP3/WAV, 1–15 s, Σ ≤ 15 s — Wan 3.0; MP3/WAV, ≤ 15 s — Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5. Audio needs at least one reference image next to it, except on Seedance 2.5. Reference videos (up to 3) are accepted by Seedance 2.5 only.
  • Priced like any clip of the same model, duration and resolution — references add nothing.
Reference mode
Removing a chip renumbers the ones after it, but the prompt text is not rewritten — re-check your @Image tags after removing a reference.

First & Last Frame

First & Last Frame animates from a start image to an end image; the prompt describes the transition. Both frames are required — JPG, PNG or WEBP up to 20 MB each, uploaded or picked from your library.

Supported by Wan 2.7, Kling 3.0, Veo 3.1 Fast. If another model is selected when you open the tab, the page switches to Wan 2.7 and shows a toast; the model list on this tab only shows supporting models.

Same formula as Text to Video and Image to Video.

FLUX 3 Max Video

FLUX 3 Max (Black Forest Labs) makes video from text, from an image — with an optional end frame the clip finishes on — and from keyframes. Clips run 5–20 s at 720p / 1080p, in any of these aspect ratios: auto, 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, 9:16. Native audio is on by default; the reasoning switch lets the model think the scene through before rendering and does not change the price.

Keyframes live on the Frames tab: up to 10 images pinned to seconds on a timeline that spans the whole clip, on a 0.5 s grid. Drag a frame to move it; no position is required, and the first frame goes to the start by default. When you change the duration, the frames move in proportion — the first stays at the start, the last at the end.

ResolutionCredits per second now · cr/sec at 720pPlan
720p43Starter and up
1080p73Pro and up
  • Opens on Starter and up; 1080p — Pro. On Free, or below a resolution's plan, any credit pack opens the model and every resolution.
  • A clip takes up to 15 min; if it fails, the credits come back automatically.

3D Camera

H3 Max · 3D Camera (MiniMax) turns one image into a video where the moment is frozen and only the camera moves — around the subject, closer, higher. It runs on the 3D Camera tab of the Video page: 1–15 s at 480p / 768p / 1080p. The frame comes from the image, so there is no aspect ratio to pick, and the clip has no sound.

The camera path is a set of points on a timeline that spans the clip (up to 20). Each point sets the orbit (degrees around the subject; full turns add up, at most 32 turns along the whole path), the elevation (-90…+90°) and the distance (0.2 close-up … 1 wide). Start from a preset, then drag or edit the points — any change switches the preset to Custom.

The prompt is optional. Leave it empty and the scene stays frozen — only the camera moves; describe the move to steer it. Prompt enhance (Off / Fast / Quality) lets the model expand your text and does not change the price.

Camera moves

Orbit 360°
A full circle around the subject at a slightly raised eye level.
Half orbit
Half a circle — from the front round to the back.
Dolly in
A push from a wide shot to a close-up, with a slight arc for depth.
Crane up
A rise from eye level to a high angle while pulling back a little.
Low fly-by
A low pass sweeping past the subject from one side to the other.
Spiral up
One full turn while climbing and pulling back.
Custom
Your own path: add, drag and edit the points on the timeline.
ResolutionCredits per second · cr/sec at 720p6 s clipPlan
480p1388Starter and up
768p20134Starter and up
1080p40268Pro and up
  • A clip is billed as its length + 0.7 s, because the model renders up to 0.7 s more than asked: 6 s at 768p = 134 credits. The price on the Generate button is what is charged.
  • Opens on Starter and up; 1080p — Pro. On Free, or below a resolution's plan, any credit pack opens the model and every resolution.
  • A clip takes up to 10 min; if it fails, the credits come back automatically.

Preset catalog

The preset catalog (sidebar → Inspiration → Presets, or the Presets pill on the Video page) is a set of ready recipes maintained by the GOGUL team: a prompt template, generation settings and input slots. Fill the slots, pick a model and aspect ratio where the recipe allows it, and press Generate.

  • A slot takes an image, a video, an audio track, a line of text or a pick from a gallery of preset images. Each accepts an upload or an item From Library; required slots must be filled before Generate is enabled.
  • The card shows the price computed from the recipe's default model and duration; the Generate button recomputes it for the model you chose. Same video formula — presets carry no surcharge.
  • Filter by model, category or title. Categories and the recipes themselves are set by the team; if a recipe changes while you have it open, the page asks you to reload it.

Limits & waiting

A few practical constraints to know before you queue something up:

  • Uploaded images must be JPG, PNG, or WEBP, up to 20 MB each.
  • One generation runs at a time — wait for the current clip to finish before starting the next.
  • Clips on Wan 2.7, Kling 3.0, Veo 3.1 Fast are usually ready within a few minutes.
  • Vidu Q4, Wan 3.0, Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5, FLUX 3 Max, H3 Max · 3D Camera run through a queue and can take up to 25 minutes (Vidu Q4 — 25; Seedance 2.5 — 20; Wan 3.0, FLUX 3 Max — 15; Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, H3 Max · 3D Camera — 10). You can leave the page — generation continues in the background and the result appears in Recent.
  • Your draft (prompt, settings, attached images) is autosaved, so nothing is lost if you navigate away.

Motion Transfer

Motion Transfer has its own tab on the Video page. Upload a photo of your character and a short video of a person moving — the Dreamactor M2 model makes your character perform that movement.

20 credits per second of the movement video, rounded up. The server measures the real duration from the file; the estimate on the button uses your browser's reading and normally matches. Nothing is charged if the video can't be read or fails the content check.

Requires Pro or higher — or any credit pack purchase, which unlocks it on every plan. The tab itself is open to everyone; the check happens when you press Generate.

Inputs

  • Character photo: JPEG or PNG up to 4.7 MB. Movement video: MP4, MOV or WebM up to 50 MB, 1–60 seconds long. An optional text note is stored with the generation but does not steer the model.
  • Before anything is charged, the photo, the note and three frames sampled from the video go through the content check.
  • Generation runs while you wait — usually a few minutes, with a staged progress view. On failure or timeout the credits are refunded automatically. The result appears in the preview panel and in Recent.

Video FAQ

Video tips

Describe motion, not a still scene
Video prompts work best when they specify camera movement, character action, and atmosphere — “slow dolly in as she turns toward the window, dust drifting in the light” beats a static scene description.
Draft at 480p, finish at 1080p
480p costs a fraction of the 720p price (×0.5 multiplier). Iterate on prompt and motion cheaply, then re-run the final version at high resolution.
Turn audio off when you don't need it
The Generate Audio toggle lowers the price of Kling 3.0, Veo 3.1 Fast clips. If you're adding your own soundtrack anyway, that's free savings.
Image to Video gives you more control than Text to Video
Starting from a frame locks composition, characters, and lighting — the model only has to add motion. Generate or pick a strong still first, then animate it.
Use the Reference mode for a recurring hero
For a series of shots with the same character, the Reference mode on Vidu Q4, Wan 3.0, Seedance 2.0 Pro, Seedance 2.0 Fast, Seedance 2.0 Mini, Seedance 2.5 keeps identity consistent — attach the character images once and tag them as @Image1 in every prompt.
Guide

Script to Storyboard

Turn any script into a professional visual storyboard in three steps: upload your script, review the AI scene breakdown, and generate frames. This guide covers the full workflow — from the first paste to exporting video-ready shots.

Quick Start

The whole flow in five moves:

1

Open the Script page

Paste your script (at least 50 characters) or upload a .txt file. No script yet? Use a Quick Start example.

2

Pick your settings

Choose a model (GOGUL 1.0 or GOGUL 3.0), a visual style, an aspect ratio, frames per scene, and optionally a Camera Look and characters.

3

Analyze Script (free)

AI breaks your script into scenes and shots — WIDE, MEDIUM, CLOSE-UP, and POV.

4

Review and edit the breakdown

Adjust scenes and shots, then check the credit estimate at the bottom of the page.

5

Generate

Watch progress scene by scene, then open your finished board with View Storyboard.

Step 1: Upload Your Script

Everything starts on the Script page. You bring the words; the settings below decide how they turn into images.

Adding your script

  • Paste directly into the editor — minimum 50 characters.
  • Or upload a .txt file up to 100 KB (up to 15,000 characters).
  • Standard screenplay format (INT./EXT. scene headings) works best, but plain prose is fine too.
  • No script handy? The Quick Start examples load a ready-made sample with one click.

Choosing a model

ModelPer frameBest for
GOGUL 1.010Fast drafts — Quick, affordable frames — ideal for first passes, layout checks, and iterating on scene structure.
GOGUL 3.020Character Lock — Premium quality with Character Lock: attached characters keep the same face, hair, and outfit across every frame.

Style, format & framing

Styles
Cinematic, Anime, Realistic, Noir, Watercolor, Illustration, 3D.
Formats
16:9 (film), 1:1 (square), 9:16 (vertical / Reels), 4:3 (classic).
Frames per scene
Choose how many frames to generate for each scene.
Camera Look
Optional presets that give the whole board a consistent lens-and-light language.

Attaching characters

  • Attach up to 5 characters per storyboard.
  • Pick them from your Character Library, or upload a photo (up to 5 MB) right on the page.
  • The AI analysis detects which characters appear in each scene and writes them into the shot descriptions. Reference photos are applied across the whole board, so each character stays consistent wherever they show up.
Character Lock works only on GOGUL 3.0
You can attach characters on GOGUL 1.0, but they won't stay visually consistent across frames. If consistent characters matter for your board, switch to GOGUL 3.0 before generating.

Analyze Script

When your script and settings are ready, hit Analyze Script — it's free. AI reads the script and splits it into scenes, each with a set of shots typed as WIDE, MEDIUM, CLOSE-UP, or POV. Nothing is generated yet — analysis only produces the editable breakdown for the next step, and you pay only for the frames you generate afterwards.

GOGUL Script page with a screenplay pasted in and AI-detected scenes ready for storyboard generation
GOGUL Script page with a screenplay pasted in and AI-detected scenes ready for storyboard generation

Step 2: Review & Edit Scenes

The breakdown is fully editable before you spend a single generation credit. This is the cheapest place to fix problems — change text here, not pixels later.

Edit everything
Scene titles, descriptions, and every shot's text can be rewritten in place.
Add & remove shots
Rebalance coverage scene by scene — from 1 to 6 shots per scene.
Shot types
Each shot is tagged WIDE, MEDIUM, CLOSE-UP, or POV, and you can change the type.
Video prompts included
Every shot ships with ready-made video prompts for Runway and Kling — useful later when you animate frames.
Cost estimate
A live credit estimate at the bottom of the page shows exactly what generation will cost before you commit.

Step 3: Generate

Hit Generate and the platform renders your board scene by scene.

  • A progress view tracks each scene as its frames are generated.
  • If any individual frame fails, its credits are automatically refunded — you only pay for frames that actually render.
  • When generation finishes, View Storyboard takes you straight to the editor with your full board.
Failed frames are auto-refunded
Occasional generation failures happen with any AI model. You don't need to contact support — credits for failed frames return to your balance automatically.

Storyboard editor

Generation is the beginning, not the end. Every frame in the storyboard editor can be fixed, reframed, extended, and upscaled.

Fix a frame

Regenerate
10 / 20
Re-roll any frame on the same model (10 credits on GOGUL 1.0, 20 on GOGUL 3.0).
Quick Edit
10
One-click preset chips for common fixes: remove an object, change the background, add detail, change the mood, fix the face, change the outfit.
Change Angle
10
Reshoot the same moment from a different camera angle.
Adjust Camera
10
Full camera control: angle, movement, lens, depth of field and lighting presets.

Extend the story

AI Fill Between
10 / 20 per frame
Generates transitional frames between two existing frames. You're charged upfront for two frames at your model's per-frame rate (10 / 20 credits each); if fewer are produced, the difference is refunded automatically.
Continue Story
50 / 90
Adds up to 4 new frames: 10 credits for the continuation plus 4 frames at your model's rate — 50 credits total on GOGUL 1.0, 90 on GOGUL 3.0. Failed frames are auto-refunded.

Coverage & quality

8-Angle Pack
200
One click produces 8 camera angles of a single frame — instant coverage for any key moment.
Upscale HD
10 / 15
Upscale any frame to 2K for 10 credits, or to 4K for 15.

Export

ZIP
Starter and up
Download all frames as a ZIP archive.
HD ZIP
Pro and up
Full board in HD quality.
PDF
Starter and up
Professional storyboard PDF layout with scene descriptions.

Manual storyboard

The Storyboard page (sidebar → Create → Storyboard) is a frame-by-frame canvas: frames on the left, the selected frame in the middle, its settings on the right. Write each frame's prompt yourself, or let AI Director draft the whole board from a short story.

Frames
Add, remove and reorder frames — up to 20 in the editor. Each frame has a prompt (with a free Enhance button), a camera block (angle, movement, lens, lighting), a frame reference (upload, From Library or Street View) and access to the board-wide characters (up to 5) and style reference. The draft is autosaved in this browser; reopening the page offers to restore it.
Generating
Generate one frame from its card, or Generate All: the board is saved to Projects and rendered in the background at 10 credits per frame on GOGUL 1.0 or 20 on GOGUL 3.0, charged upfront after a confirmation dialog. Failed frames are refunded. Generate All is capped by your plan's frame limit (see Plans); the editor lets you add more, so the check happens when you press the button.
AI Director
The AI Director button above the Overall story field reads your story and replaces the board with a directed set of frames: a prompt, camera angle, movement, lens and lighting for each. Free. It asks for confirmation when frames already have prompts or images, and drafts no more frames than your plan allows.
Google Street View as a reference
Open Location in the right panel, search a place and drop the marker; Use as Reference attaches the Street View still as that frame's reference. Free, no plan gate; it replaces any other reference on the frame. Some spots have no Street View coverage.
Public link
Share in the storyboard editor creates a public link valid for 30 days; sharing again extends it by the same span and keeps the same address. Viewers see the frames, descriptions, camera chips and copyable video prompts. A switch in the same dialog turns the link off. Shared pages are not indexed by search engines.
Projects
Projects (sidebar → Library → Projects) lists your storyboards with search, a style filter and sorting. A card can be opened, duplicated or deleted; renaming happens in the storyboard editor.

From storyboard to video

There are two ways to put your board in motion:

Path A — Copy the ready-made prompts
Every shot has a Video tab with prompts pre-written for Runway Gen-3 and Kling. Copy a prompt, paste it into your video tool of choice, and use the frame as the visual reference.
Path B — Animate inside GOGUL
Download a frame and bring it to the Video page as a Start Frame. Vidu Q4 is available on every plan; the other models unlock at their plan tier or with any credit pack purchase — the exception is Veo 3.1 Fast, which stays plan-exclusive. See the video models table for durations, inputs and prices. Full video guide →

Storyboard pricing

Everything in the Script to Storyboard flow is priced in credits:

ActionCredits
Analyze scriptFree
Frame — GOGUL 1.010
Frame — GOGUL 3.0 (Character Lock)20
Regenerate frame10 / 20
Quick Edit10
Change Angle / Adjust Camera10
AI Fill Between (per frame)10 / 20
8-Angle Pack200
Continue Story50 / 90
Upscale 2K / 4K10 / 15
PDF export (Starter+)Free

Example estimate

ActionCredits
Script analysisFree
4 scenes × 3 frames on GOGUL 1.0 (12 frames × 10)120
Total120

Storyboard FAQ

Storyboard tips

Write for the camera
Use scene headings (INT./EXT. LOCATION – TIME) and describe what the camera sees — actions, light, and setting — not internal thoughts. Cleaner scene detection means a better breakdown.
Draft on GOGUL 1.0, finish on GOGUL 3.0
Iterate on structure with cheap GOGUL 1.0 frames, then regenerate the final board on GOGUL 3.0 once the scenes and shots are locked.
Set up characters before generating
Attach characters with clear, well-lit photos and switch to GOGUL 3.0 so Character Lock keeps them consistent. Don't mix models within one board.
Fix text, not pixels
The Review step is free to edit. Rewriting a shot description costs nothing; regenerating a frame costs credits. Get the words right first.
Use Camera Look for a unified board
A Camera Look preset applies one lens-and-light language to every frame, so the board reads as a single production rather than a collage.

Character library

The Characters page (sidebar → Library → Characters) stores the people who appear in your storyboards. Attach a saved character to a script and Character Lock keeps the same face, hair and outfit in every frame.

  • Create a character from a photo, or save any generated image as a character
  • AI analysis detects facial features, hair, clothing and body type — free
  • Generate an AI portrait for 10 credits
  • Generate Character Views — 4 extra angles of the same person for 40 credits, charged upfront; any angle that fails is refunded
  • Reuse a character across any number of storyboards

Library size by plan

Free — 1 (20 with a credit pack)Starter — 5 (20 with a credit pack)Pro — 20Studio — 999Agency — 999

At the limit the create button opens an upgrade dialog. Any credit-pack purchase raises the limit to 20 on plans with a lower limit.

GOGUL new character page with photo upload and AI analysis of facial features, hair, and clothing
GOGUL new character page with photo upload and AI analysis of facial features, hair, and clothing

Multi-Angle

Multi-Angle (sidebar → Tools → Multi-Angle) generates the camera angles you pick from a single image — each angle as its own full-size frame in the format you choose. The camera is set to an exact angle (horizontal, vertical, distance), as on the Control tab, so the subject and the scene stay the same.

Available from the Starter plan, or on any plan after a credit-pack purchase. Each angle costs 25 credits and is charged separately; if an angle fails, its credits are returned. Regenerating a single angle costs the same. The total is shown on the page before you generate.

Two modes

  • Character Turnaround — white background, the same person from different angles
  • Scene Angles — the same scene from different camera positions

Preset angles

Front ViewBack ViewLeft SideRight SideClose-UpBird's EyeLow Angle3/4 View

Or describe any custom angle in words.

Burned-in subtitles

The Subtitles tab (sidebar → Create → Subtitles & Dubbing → Subtitles) transcribes the speech in a video and burns styled captions into it. Pick a source, a caption style and, if needed, the spoken and target languages; the price appears as soon as the video is read, and the finished MP4 lands in the result card and the History list on the same page.

How it works

1

Pick a source

Upload a video or choose one of your generated videos under My videos.

2

Choose a style

The style strip previews each caption look; static and animated styles are priced differently.

3

Languages and dictionary

Leave the spoken language on auto-detect or set it; pick a translation language if the captions should be in another language; add brand names to the dictionary.

4

Check the price

The right column shows the billed minutes, the rate and the multipliers before you press Generate.

5

Render and download

Rendering takes a few minutes and runs while you wait. Download from the result card or later from History.

Sources and limits

Accepted: MP4, MOV, WebM, M4V and GIF, up to 60 MB and 10 minutes. Generated videos become available here once they are stored in your library. The server re-measures the file, so the final price is based on the real duration and resolution.

Have captions already? Paste an SRT file instead of transcribing: the captions are burned in as given, and the dictionary is ignored.

Pricing

Billed per started minute of video (a shorter clip counts as one minute):

  • Static caption style — 34 credits per minute
  • Animated caption style — ×2 the static rate
  • Video above 1080p — ×2
  • Translation into another language — +67 credits per minute

21 static and 9 animated styles to choose from; animated ones bill at ×2.

Fine-tuning (font, colors, position, shadow) applies to static styles only.

Source video up to 10 minutes and 60 MB. No plan gate — every plan can use Subtitles.

Example
A 3-minute 1080p video with an animated style and translation costs 405 credits.
Translation
Set Translate to for captions in another language. Translation replaces the original captions — the output carries one language, and no SRT file is exported. It adds a flat per-minute fee on top of the style rate.
Dictionary
Rows of the correct spelling plus the misheard variants, comma-separated. The list is sent with the transcription as a hint for names and brand terms. It is capped — Add word disables at the cap — and ignored when you supply your own SRT.
Output
Always an MP4 with the captions rendered in. It is listed only in the History section of the Subtitles page — not in Projects or the media library — and can be downloaded from there.
What happens if a generation fails?
If the render fails after the charge, the credits come back automatically. Rejections before the charge (wrong format, too long, too big, unreadable file) cost nothing.

Dubbing

The Dubbing tab (sidebar → Create → Subtitles & Dubbing → Dubbing) translates the speech in a video into another language in the speaker's own voice, with lip-sync. Upload a video or pick one of yours, choose the languages and the quality; the price appears as soon as the video is read, and the finished dub lands in the result card and in Recent dubs.

How it works

1

Pick a source

Upload a video with speech or choose one of your generated videos under My videos.

2

Choose the languages

Leave the spoken language on auto-detect or set it, then pick the language to dub into — search by name, native name or code.

3

Choose the quality

Standard or Pro. In Pro you can also set the voice match and a regional accent.

4

Keep words untranslated

Optionally list brand, product and people names that must stay as they are.

5

Confirm and start

Confirm you own the rights to the video and the voices in it, then press Dub video. A dub usually takes 2–5 minutes; you can close the page meanwhile.

Standard and Pro

  • Standard — 28 credits per second of the source video, 86 languages, cloned voice and lip-sync. From the Starter plan or after any credit-pack purchase.
  • Pro — 45 credits per second, 94 languages plus 14 regional accents, voice match from 0 to 10 (default 7), several speakers, background music kept. From the Pro plan or after any credit-pack purchase.

Pricing

The price is the length of the source video times the per-second rate of the quality; every started second counts. The length is measured on the server before anything is charged, and the button shows the exact price.

Example
A 42-second video costs 1176 credits on Standard and 1890 credits on Pro.

Up to 10 minutes and 300 MB per video, MP4, MOV or WebM. The video must contain speech — a silent file is refused before anything is charged.

How long it takes
Usually 2–5 minutes: most of it is the provider's queue, the dub itself is quick. You can close the page — the result appears in Recent dubs.
Keep these words
Up to 1000 names that must not be translated — a brand, a product, a person. Each entry is a short phrase.
Rights to the voice
Dubbing clones the speaker's voice, so you confirm that you own the rights to the video and the voices in it before every dub.
Original video
An uploaded original is kept for 7 days after its last dub, so you can dub it into another language; dubbed videos are kept like any other generation. After that, upload the original again.
Subtitles for a dub
Add subtitles sends the finished dub to the Subtitles tab with the dub's language already selected; subtitles are priced separately.
What happens if a generation fails?
If a dub fails, the credits return to your balance automatically.

Avatar (Live)

Avatar (sidebar → Create → Avatar; the page is titled Avatar Studio) turns a portrait photo and an audio track into a talking-head video: lip-sync, head movement and expression follow the audio. Options: a short direction note, 720p or 1080p output, and a fast mode.

25 credits per second of audio, with a minimum of 150 credits per generation. The duration is read from the audio file in your browser, and the price is shown under the settings before you press Generate.

Requires Pro or higher. A credit pack does not unlock it — the gate is plan-only, checked on the server when you press Generate.

Requirements
Photo: JPEG or PNG, one clear face; the size limit is shown on the upload zone. Audio: MP3, uploaded or picked from your media library — there is no in-browser recording or text-to-speech on this page. A file of another type is ignored silently, so if nothing happens after a drop, check the format.
Waiting
Generation runs in the background: the page polls for the result and shows a time estimate based on the audio length — expect several minutes for a short clip, longer at 1080p. You can leave and come back; the active task is remembered in this browser.
Avatar (Live)
Finished videos stay in the History strip on the page: download, select and delete them there. A failed generation is refunded automatically.

Community

Community (sidebar → Inspiration → Community) is a public feed of images and videos made in GOGUL. Browsing, publishing, liking and remixing are free on every plan.

Feed
One grid with filters — All / Images / Videos, style and model chips, popular tags — and a sort: Newest, Trending, Most Liked, Most Discussed. Trending weighs likes, comments and remixes against age. Above the grid: Trending Now (last week) and Recommended for You (based on the tags of works you liked).
Publishing
Publish from an image card, the storyboard editor or the Video page. You pick up to 5 tags, whether the work shows on your profile, and whether to submit it to the feed. Submitted works are reviewed before they appear unless the team has switched on auto-approval; profile-only works skip the review. A published work cannot be taken down by the user — contact support to remove one.
Likes, download, remix
Like toggles a heart and notifies the author. Download saves the original file, without a watermark. Remix opens the Image or Video page with the work's prompt, model and style prefilled (style is not carried over for videos); the author sees a remix count and gets a notification, but your new work carries no link back to the original.
Collections
Collections are boards of works. Create one from the bookmark menu on any work (public by default), rename it or make it private later, add or remove works. Public collections show on your profile; the number of collections per user is capped.
Following authors
Follow from an author's profile (tap the avatar on any card). The profile shows works, followers, following, likes, views, remixes and badges, with tabs for images, videos, collections and liked works. There is no separate following feed — follows drive notifications and profile visits.
Comments
Comments with one level of replies, likes on comments and a Translate button. You can delete your own comments; the length cap is shown by the counter under the field.
Badges
Awarded automatically as you publish (Creator), get remixed (Trendsetter), gain followers (Social) or fill collections (Collector); special badges are granted by the team. They show on your profile with progress toward the next tier and arrive as a notification.
Search
The search box looks through the prompts, titles, descriptions, styles and models of works, the names and bios of creators, and tags. Results drop down as you type; Enter applies the query to the grid.

AI assistants

GOGUL works right inside Claude. Once connected, Claude creates images and videos from your description, shows your credit balance and checks whether a generation is ready — without leaving the chat.

You add GOGUL in Claude as a custom connector and sign in with your GOGUL account. Generations spend the credits of your GOGUL account at the usual prices. To disconnect, use Claude's connector settings or Settings → Connected apps in GOGUL. The step-by-step guide is at /integrations/claude.

Plans & limits

Plans renew monthly credits and set the limits below. The Free plan is a one-time balance.

PlanCreditsPriceFrames / storyboardCharactersVideo modelsExports
Free100Free8 (12 with a credit pack)11ZIP: No · HD ZIP: No · PDF: No
Starter1,000$9/mo$7/mo billed yearly ($86)1259ZIP: Yes · HD ZIP: No · PDF: Yes
Pro3,500$29/mo$23/mo billed yearly ($278)202010ZIP: Yes · HD ZIP: Yes · PDF: Yes
Studio9,000$59/mo$47/mo billed yearly ($566)99999911ZIP: Yes · HD ZIP: Yes · PDF: Yes
Agency30,000$169/mo$135/mo billed yearly ($1620)99999911ZIP: Yes · HD ZIP: Yes · PDF: Yes
  • Monthly subscription credits do not roll over. Purchased credits never expire while your account is active.
  • Upgrade or cancel any time on the Billing page — you keep access until the end of the paid period.
  • Card payments through Stripe; Click and Payme are available as local payment methods — see Payment methods below.

Credit costs

What each action costs. Video and subtitle prices are formula-based — see their sections.

ActionCredits
Storyboard frame — GOGUL 1.010
Storyboard frame — GOGUL 3.0 (Character Lock)20
Regenerate frame10
Continue Story (plus frames at the model rate)10
Change Angle on a frame10
Adjust Camera (one angle)25
All-angles pack200
Edit image (1K)10
Edit image (2K)20
Brush / inpaint15
Outpaint12
Relight12
Remove background5
Upscale 2K10
Upscale 4K15
AI character portrait10
Character Views (4 angles)40
Multi-Angle, per angle25
Subtitles, per minute (static style)34
Dubbing Standard, per second of video28
Dubbing Pro, per second of video45

Credit packs

Need more credits? Buy a pack any time, on any plan.

100 credits — $3
500 credits — $12
1,500 credits — $29
  • Purchased credits never expire while your account is active.
  • Any pack purchase also unlocks plan-gated models and quality tiers, including 2K in the editor (except plan-exclusive video models); removes the watermark; opens Multi-Angle, HD frames, HD download and voice cloning; and lifts the storyboard frame ceiling to 12 frames and the character limit to 20.

Payment methods

Card payments run through Stripe: subscriptions renew monthly or yearly, credit packs are one-time. Two local methods appear on the Billing page for Uzbekistan — detected by location, profile country or the Uzbek interface language — and are priced in so'm:

  • Click — credit packs only. Press Pay via Click on a pack card, finish on Click's page, and the credits arrive once Click confirms the payment.
  • Payme — subscriptions only. Press Subscribe via Payme on a plan card. The plan is activated for a month after payment; there is no automatic renewal, so pay again to extend.

Referral rewards are granted for card subscriptions only.

Referrals

Share your link from the Referrals page (sidebar → Account → Referrals). Your code is attached when a friend opens the link and registers.

When the friend pays for a subscription by card, you receive 200 credits and they receive 200. Nothing is awarded at signup, and credit-pack or local-method payments do not count.

Referrer rewards stop at 5,000 credits in total (25 paid referrals); the friend's bonus is always paid.

Self-referrals and referrals from the same IP address or the same email are blocked. Rewards are reversed if the subscription is cancelled or refunded.

Milestone bonuses

5 paid referrals — +300 credits, once10 paid referrals — +750 credits, once25 paid referrals — +2,000 credits, once

How it works

1

Copy your link

The Referrals page shows your personal link and share buttons.

2

A friend signs up

The referral is recorded as pending when they register through your link.

3

They subscribe

The first card payment for a plan turns the referral into a reward for both of you, and the page updates its stats.

Settings & security

Settings (sidebar → Account → Settings) has four blocks: profile, subscription, security and data.

  • Profile: display name, avatar and interface language. The email address cannot be changed here.
  • Change password: enter the current password, then the new one twice.
  • Two-factor authentication with an authenticator app (TOTP): scan the QR code or type the key, then confirm with a code. There are no backup codes, so keep the authenticator safe. Disable it from the same block.
  • Export your data downloads a JSON file with your profile summary and all storyboards with their frames. Generated images, characters and videos are not included — download them from their own pages.
  • Delete account is immediate and irreversible: type DELETE to confirm. Files, storyboards, generations, community posts and comments are removed, and a Stripe subscription is cancelled without a refund for the remaining period. A short anonymized payment record is kept where payment providers require it.

Upload limits

Limits enforced by the product configuration:

  • Video prompt — up to 7,000 characters
  • Reference images per generation — up to 16, depending on the model
  • Subtitle source video — up to 60 MB and 10 minutes
  • Dubbing source video — up to 300 MB and 10 minutes

Per-page upload sizes (character photos, storyboard references, Multi-Angle input) are shown next to each upload button.

Contact support

  • Email: [email protected]
  • Response time: within one business day
  • For generation issues, include the storyboard name and the frame number.

Credits & billing

Models & quality

General questions