Skip to main content

Create and Edit in Plain Language

Summary: The Workspace AI agent knows every XBRUSH capability at the API level. Say what you want in plain language and it works out the steps, picks the tool and engine, writes the prompt, and runs it for you. You do not have to know the product to use it. Each piece of work is a session; sessions are organized in folders, every session can be worked on with your team, and Quick start, LARP, and Workflow recipes give you ready-made sessions to copy.


Two Ways to Work, One Screen​

Workspace home with the greeting, the chat box (Chat · Direct · Attach), Recent sessions, and the New folder card

Workspace home — a greeting, the agent's chat box with Chat · Direct · Attach on the left and the Send · New line · @ Pick character hints on the right, Recent sessions, and a New folder card. The LARP row starts below; scroll further for Quick start, Workflow recipes, and All sessions. The manual controls open from the Direct button in the chat box.

Workspace gives you both approaches at once:

  • Chat mode in the input box — describe the result you want, and the AI agent does the rest.
  • Direct mode in the input box — click Direct at the bottom left of the input box and a panel opens inside it, with Image / Video / Audio tabs and their sub-tabs (for images: Generate · Edit · Outpaint · Remove BG · Upscale · Layer Split), engine selection, aspect ratio, Count, and resolution, all under your direct control.

They are not separate modes. Ask the agent to do something, then adjust the details with Direct; or set everything yourself and ask the agent only for the awkward step. This page is about the first approach. For the second, the existing guides cover every control in detail — see Working the manual way at the end.


Say It, and It Runs​

A session in progress: the user's request, the agent's plan with engine and estimated cost, and a video result card tagged Generate #4

A session in progress. The user's request (right-hand bubble), the agent's reply stating its plan, engine, and estimated cost, and the result card flow like a conversation; each card is tagged Generate or Edit with a number (here a video, Generate #4). The chat box stays pinned at the bottom.

This is what makes the agent different from a prompt box. Take a session that turned one character image into a 3×3 sprite sheet, and follow what actually happened:

ℹ️ This example was recorded in an earlier version, when the agent used GPT Image 2 (gpt-image-2-edit). The current default image engine is GPT Image 2.5, so the engine name and cost you see today may differ. The way the agent plans and runs the work is the same.

1. The request was ordinary language. The user attached a character image and wrote, in Korean: "I want to turn this cat archer's shooting scene into a 2D sprite for a game. Make me a 9-frame 2D sprite." No tool named, no engine chosen, no format specified.

2. The agent stated its plan and the cost first. "I'll build a 3×3 nine-frame archery sprite sheet from the attached cat archer. gpt-image-2-edit, low quality, ~2.4C." You see what it intends to do and roughly what it will cost before anything is spent.

3. The agent executed it as you. The entry in the stream shows Edit — not Generate — with the engine GPT Image 2, the reference image, and a detailed English prompt the agent wrote itself:

Turn this character into a 3×3 grid sprite sheet (9 frames) of an archery shooting animation: drawing the bow, aiming, releasing, and follow-through. Keep the same 2D game character design, consistent size and side view in every frame, flat clean plain background, evenly spaced cells.

4. It reported the result and offered the next step. "The 9-frame (3×3) archery sprite sheet is done (2.4C charged). Would you like me to slice it into individual frame files?"

Notice what the user never had to know: that this is an Edit job rather than a Generate one, that GPT Image 2 handles grid layouts well, how to phrase a sprite-sheet prompt in English, or what "consistent size and side view in every frame" would prevent. Because the agent knows every XBRUSH API, it translates a requirement into the procedure and the calls that satisfy it.

That applies across images, video, and audio alike. Generation, editing, background removal, upscaling, outpaint, voice, music — if XBRUSH can do it, the agent can be asked for it in plain language.


Starting a Session​

A new, empty session tab reading "Start anything with a chat"

A new session opens in its own New session tab and waits at "Start anything with a chat", with a Drop files here zone ("ready to use as material").

In Workspace, one piece of work — a conversation with the agent and everything it produced — is called a session. A few buttons, such as Create task on a recipe, still say "task"; a task and a session are the same thing, and this guide says "session" throughout.

Click New session under Quick start or + in the tab bar, or just type in the chat box on the home screen. Ways to begin:

  • Type an instruction. Describe the outcome, not the operation. The placeholder rotates through examples of how loose you can be: "Just say where it will be used and we'll match the format — 'vertical, for an Instagram story'.", "Refine as you go — 'make that one a bit brighter'."
  • Drop a file. Drop files here — ready to use as material. Anything you drag in becomes source material immediately.
  • Call a character with @. Type @ in the input box to pick a character from your character library — the hint @ Pick character is shown at the right of the box. To add a character to the library, use the New character Quick start, which makes a character sheet and saves it to the library. Once it is there, you can bring the same character into any session by name.

You can keep talking after the first result. The conversation is the session, so follow-ups like "make the background lighter" or "now turn it into a video" continue from what is already there.


Folders, Sessions, and Your Team​

Results are not dumped into one pile. Every result belongs to a session, and sessions live in folders — the breadcrumb under the tab bar (e.g. Home) shows where you are, the folder tree icon at the far left of the tab bar moves between folders, and the home screen shows Recent sessions, a New folder card, and, at the bottom, All sessions.

Create folders per project, client, or campaign and move sessions into them, so a month later you are opening a folder rather than scrolling a timeline.

Every session can be worked on with your team. Teammates open the same session, add their own instructions, and see results appear in the same stream — the conversation shows who ran what. There is no exporting and re-sharing between people; the session itself is the shared workspace.

  • For folder management, permissions, and result-level actions, see the Workspace guide.

Copy a Ready-Made Session​

The Quick start · 61 row and the Workflow recipes · 41 row on the Workspace home

The Quick start · 61 row (New session, New character, Edit by Chat, LARP, Remove Background, Studio Product Shot, Create an Image, Make a trending video) with See all quick starts, followed by the Workflow recipes · 41 row, where each card shows its steps, image count, and credit range, with See all recipes. The LARP row sits above these; see LARP below.

If you do not yet know what to ask for, start from something that already works. The home screen carries these rows:

RowWhat it is
Quick startSingle jobs that finish in one instruction — for example New session, New character, Edit by Chat, LARP, Remove Background, Studio Product Shot, Create an Image, Make a trending video, and more
LARPReference videos you can remake with your own character as the lead — pick one and upload a photo
Workflow recipesMulti-step packages that produce a whole set — each card shows its step count, the number of images it produces, and a credit range

See all quick starts and See all recipes below each row open the full list inside Workspace — 61 Quick start entries and 41 recipes at the time of writing.

Can't find Recipes or Quick Start in the top menu?​

Recipes and Quick Start appear in the top menu only while you are logged out. Once you log in they drop out of the header, so use one of these entry points instead:

Entry pointWhereOpens
See all quick starts · See all recipesBelow each row on the Workspace homeThe full list inside Workspace
Quick Start · AI Recipes linksResources in the footer at the bottom of every pageThe public pages (/en/quick-start, /en/recipes)

Public Quick Start page while logged out

The public Quick Start page ("Upload. Say one line. Done."), opened from Quick Start in the logged-out header (Open Studio · Workspace · IP Market · Social Media · Recipes · Quick Start · API · Creator Partners). Each card shows a before/after example, the estimated credits per run, and the free runs you get on signup — e.g. Replace Image Text, Brighten Photos, Marketplace Cutout, Upscale, Blur Faces, Product Scenes.

Public AI Recipes page while logged out

The public AI Recipes page ("Upload once. Get the finished set."), opened from Recipes in the logged-out header. Recipes are grouped into categories — Commerce, Ads & Promo, Social & Channel, Branding, Store & Space, People, Character & Art, Occasions — and each card shows how many images it produces, the estimated credits, and "Drafts covered free".

On a recipe's public detail page, Open this recipe opens it straight in your Workspace. The public pages feature a selection only — log in and use the Workspace home for the full list.

Quick start — one instruction, one result​

Edit by Chat

Edit by Chat, with a before/after example at the top.

Open a Quick start card and it explains itself with a worked example, then asks only for what it needs. Edit by Chat shows a plant photo whose background became a beach, and puts it plainly: "Upload an image and write the change in one line. 'Make the background a beach', 'remove that person' — words are the tool."

Drop or paste an image (PNG · JPG · WebP), write the change in the box — "How should we change it? — e.g. make the background a beach" — and press Run.

LARP — same video, your character as the lead​

The LARP · 6 row on the Workspace home: six reference video cards and the Latest LARP button

The LARP · 6 — Pick one, add your photo row on the Workspace home. Each card is a reference video to recreate; Latest LARP opens the dedicated LARP page.

LARP remakes an existing video with your character in the lead role — same shots, same motion, a different star. This feature used to be called Recast; the interface now shows LARP everywhere. The Workspace home has a LARP · 6 — Pick one, add your photo row (6 reference videos at the time of writing) that tells you exactly what to do.

  • Reference cards (Recreate) — each card is a reference video you can follow. Click Recreate on a card and upload a photo of your character (or the person you want in the lead).
  • Latest LARP — opens the dedicated LARP page (xbrush.ai/en/larp).
  • Direct link — https://xbrush.ai/workspace?open=video-recast opens LARP straight away when you are logged in.

The LARP page: a before/after comparison video with a drag handle, the "One take, two leads" text with a Recreate button, and a LARP button at the top right

The LARP page ("Same video, your character"). Drag the handle on the comparison at the top to compare the source with the result. Below it, the cards are split into References and Community tabs, with Show more at the end. The homepage (xbrush.ai/en) also has a LARP row and a Latest LARP → link.

LARP is also available as a Quick start entry and as the LARP sub-tab of the Video tab in the Direct panel — see the Workspace guide.

Workflow recipes — a full set from one setup​

Space Promo Kit

Space Promo Kit — the core steps produce 15 images; optional steps add more.

A recipe is a whole deliverable, not a single image. Opening one shows you everything before you commit:

  • The steps and what each produces — Space Promo Kit's core steps are 3 tone directions, 2 hero-shot retouches, 2 virtual stylings, 1 info card, 5 Instagram carousel images, 1 story notice, and 1 wide cover. (Optional) steps such as tour reel clips, a reel compilation, and background music can follow. The card on the Workspace home shows 7 steps · 15 images · 36–309C.
  • Starting materials — 1–3 photos of the space · name, address, opening hours · a one-line description. As the note says, the AI will ask for these in order after you create the session. Having them ready speeds things up.
  • The expected total — 15 images from the core steps, roughly 36–309C in credits. Optional steps add more output and credits.

Other cards work the same way — for example, Store Detail Page Pack shows 11 steps · 22 images · 48–412C. The numbers on each card are estimates and can change as engines and prices are updated.

Press Create task to start a new session, and the agent walks you through it, asking for each material in turn. You supply the raw ingredients; it handles the sequence.

💡 Recipes are also the fastest way to learn what the agent can do. Run one close to your work, watch which tools and engines it chooses at each step, and you will know what to ask for directly next time.


Working the Manual Way​

Nothing here replaces the manual controls. Everything the agent does, you can do yourself from the Direct panel in the input box with full control over engine, parameters, and iteration — and for repeatable production work that is often what you want.

Those controls are documented in full already: