Chat Bar: Generating Images

The chat bar is the floating prompt composer for generating images (and video / 3D) on the board. This page covers image generation: writing prompts, picking models, and feeding in references, LoRAs, and aspect ratios.

Generate an image (text-to-image)

To make an image from just a prompt:

Make sure the category pill (top-left of the bar) shows Image. With nothing selected on the board its icon is a T (text-to-image).
Type your prompt in the input.
Press the Send button (the round arrow-up button, bottom-right) or hit Ctrl/Cmd+Enter. Plain Enter inserts a newline — it does not send.

The result appears as a node on the board.

Image-to-image

To generate from an existing image:

Select one image node on the board. The category pill icon changes to a single image (one selection) or stacked images (2+); the mode is now image-to-image.
Selected images show as thumbnails in the selection strip above the prompt.
Type your prompt and Send.

With 2 or more images selected, the @ key in the prompt switches to image references — type @image1, @image2, etc. to refer to a selected image by index.

Category selector (Image / Video / 3D)

The leftmost pill picks what the bar does. Click it to open the mode menu: the three categories, then Workflows, then Agents.

Image — stills from your prompt.
Video — animate a selected image from your prompt. There is no text-to-video: with nothing selected, the bar asks for an image. Select a clip instead and the bar switches to video-to-video, the way selecting an image switches Image to editing it; see Video to Video.
3D — turn a selected image into a mesh; the prompt is optional.
Workflows — run one of your saved workflows on the selected images. The bar turns amber and swaps the prompt for the chosen workflow, its inputs, its steps and the estimated credits; see Workflows.
Agents — under a line, with an arrow: point at it to list your agents (and New agent). Picking one turns the chat bar into that agent's panel. It is for Pro and Max plans and team plans, so a free account's menu ends at Workflows.

Picking Image, Video or 3D is also the way back from Workflows or from an agent.

Enhance is not one of the pill's choices: it is a mode of its own, locked on the chat bar inside the 3D Editor. It sends the frame the editor shows to the image models.

In the 2D editor the mode is locked to Image, and neither Workflows nor agents are offered.

Generation mode indicator

The Image pill's icon doubles as a live indicator of what will be generated:

T — text-to-image (no board selection)
single image — image-to-image with one reference
stacked images — multi-image input
dotted box — inpainting (a selection has been drawn in the Editor2D editor)

Model picker (override per prompt)

Next to the category pill is the active model pill — the model the chat will actually call, auto-suggested from your category, references, LoRAs, and image inputs.

Click the model name to override the model for this prompt (a popover groups the compatible models into a card per maker — Black Forest Labs, Google, OpenAI, … — cheapest first, with credit cost on hover).
Auto sits in its own tile at the top, lets the suggestion engine pick, and names the model it picked (or your pin).
Every model shows what it is good at under its name, and badges for what it takes: its frames ("Start + optional end"), how many images, clips and audio tracks it reads, and whether it takes a LoRA.
Below the runnable models, closed lines hold the models a different selection would run: Needs image input, Takes fewer inputs, and Not with this setup (a LoRA or a design reference). Click one to open it; each model in it says what would let it run, such as "Select an image to use" or "Select 1 more image".
Click the pin on any row to set it as your default for the current mode (text-to-image, image-to-image, multi-image, etc.). An override lasts until the generation mode changes — selecting or deselecting images, or switching category, resets it to Auto for the new mode; pins persist.

Prompt favorites & history

Docked inside the prompt field, at its bottom right:

Star (Sparkle icon) — toggles saving the current prompt to favorites; click the favorites popover to refill the input from a saved prompt. Favorites are scoped to the current mode.
Clock — prompt history; click any entry to refill the input.

Presets

The chat bar has one-click presets to fill your prompt with curated starting points. Presets exist per kind of run. The picker shows three kinds: Text to image, Image to image and Image to video. A preset saved for several images (multi-image to image) counts as Image to image there, and its tile is tagged I2I. The preset editor's Type still offers both Image to image and Multi-image to image, because a preset keeps that difference when it is saved. The picker opens on every kind, not only the one the chat bar is doing now. Where the grid mixes kinds, a chip per kind, with its count, narrows it.

To apply a preset:

Click Presets (next to the model selector).
A panel opens on Framo's presets of every kind, the ones that fit what the bar is making first. The sidebar on the left holds Framo's presets, by kind and category, and your own presets and folders under your name. A search field at the top searches both.
Click a preset to fill the input with its prompt. A preset made for another kind of run is greyed out and says which. Some Framo presets work as two kinds — a logo preset runs on a logo you attach, or on a letter its prompt names — and those show under both kinds, tagged with both.

The pill then shows the preset's name. Once you edit the prompt, it clears — a visual cue that you've drifted from the saved preset.

The same panel is where presets are managed. New preset, the first tile among your own presets, starts one from the chat bar's prompt (or an empty page when the bar is empty); the pencil on a preset's picture edits it; the bin deletes it. Folders are made, renamed and removed in the sidebar, and a preset is filed by dragging it onto a folder.

Presets are also on the board: open an image's actions menu (the arrow beside the image) and pick from Framo Presets, or from your own presets, which sit in the row above it named after you (for example "Roman's Presets").

Your own presets and reference images

Your own presets are prompts you saved. One of them can carry up to 6 reference images from your library, added in the preset editor. When you run the preset from an image's actions menu, that image is sent first, as image 1, and the references follow as image 2, 3, … in the order the preset saved them. So a preset's prompt names its images in that order, for example "Recast the subject of image 1 in the glazed clay of image 2".

A preset with reference images runs only where it can use every one of them:

From an image's actions menu on the board, on a run model that takes the image plus every reference. When the model takes fewer, the preset is greyed out with "Needs a model that takes N images"; the model presets run on is set in Settings, under Models.
Not from the chat bar's presets, where it reads "Run it from an image's menu on the board".
If you delete one of its reference images, it reads "Reference N was deleted" until you remove or replace that image in the preset editor.

Let the AI save a preset

In the refine chat (the wand), ask the AI to turn a prompt into a preset — "save this as a preset", or "make a preset that repaints any photo in the style of image 2". It saves the preset to your own presets straight away, with the attached images you name as its references, and files it in a folder: one of yours that fits, or a new one it creates (you can name the folder yourself). An image-to-image preset gets the image it transforms as its cover unless you ask for another. A line under its reply confirms where it went ("Saved to Roman's Presets › Lighting: …"). It writes the saved prompt in the order the preset will send its images, so the image you later run it on is image 1. When it is not clear which attached image is the one to transform and which are references, it asks before saving. You can edit or delete the preset afterwards like any other.

Share a saved prompt with your org

From the Prompt Favorites popover (Sparkle icon, top-right cluster):

Right-click any saved prompt card or click the organization button on hover.
Toggle the Building2 icon (org logo or building icon) to share / unshare with your organization.
Shared prompts appear in the org filter tab at the top of the favorites popover, visible to all members of your workspace.

Refine with AI (prompt enhancement)

Click the wand (Wand2) icon at the bottom right of the prompt field to open a chat-style prompt-refinement surface. You converse with the AI to improve your prompt, then accept a result to drop it back into the input. This is a Pro feature — free users are prompted to upgrade.

With several images selected, the AI numbers them in the order shown. When you ask it to transform one image using the others, the prompt it writes names each image by number and role ("restyle image 1 in the palette of image 2"); if it cannot tell which image is which, or whether you want the images named, it asks first. It can also save the result as one of your presets — see "Let the AI save a preset" above. See Sketch pipelines & prompt tools for the full enhancer.

Auto-enhancement (Enhance ON/OFF)

The Enhance toggle pill in the bottom row (scroll on it or click to flip) turns on a vision pre-pass: for image-to-image, the input image is analyzed and your prompt is reconciled before generation. Off by default.

The "+" menu: Design reference, LoRA, Look, Matrix

The + button (bottom row, left) adds optional capability panels:

Design reference

An image the model uses alongside your prompt to adopt color, composition, or structure (model-dependent).

Add via + → Design reference, then click the panel to pick an image, drag a node in, or use the eyedropper to pick from the board.
Shortcut: hold Alt and click any image/video/model node on the board to set it as the reference.
Strength (header pill, ~0.20 loose to ~1.00 close): scroll to adjust or click for a slider.

LoRA

Add via + → LoRA (up to 4). Each LoRA bubble lets you pick a trained model, set its scale, and open its details.
Picking your first LoRA on an empty prompt drops in the trigger-word phrasing as a starting point.
When a LoRA is configured, type @ in the prompt to mention its trigger word.

See Training & using LoRAs for training and trigger words.

Look (camera / lens / film)

Stacks cinematography modifiers onto your prompt. Add via + → Look.

A Combo dropdown applies curated presets (Cinematic, Epic, Portrait, Street, Dreamy, Anamorphic).
Three slots — Camera, Lens, Film — set independently. Click a slot for its picker, or wheel-scroll over it to cycle.
The dice randomizes all three.
Shares state with the AI sidebar's modifier bar. Warns if your prompt already mentions camera/lens/film info.

Matrix (variations grid) — image only

Generates a grid of variations in one call. Add via + → Matrix.

Composition: 2×2, side-by-side, stacked, triple side-by-side, triple stacked.
Creativity: Subtle, Balanced, Wild.

Aspect ratio

The ratio pill (bottom row) sets output width-to-height. Click for a grid of ratios, or scroll on the pill to cycle. The picker only shows ratios the resolved model supports, and may snap to the model's default when you switch models. Auto (when offered) matches the input image's dimensions. Hidden in the Editor2D editor (output is sized from the canvas).

Number of images & parameter variation

The N image(s) pill (bottom row, Pro/admin only) sets how many images per prompt (1–4). Scroll to adjust or pick in the popover. Cost scales linearly.

When set to 2 or more, a Vary across batch section appears: click the eyedropper, then click any varyable setting pill in the bottom row to vary it; set a ± range and the values are spread evenly across the batch (one image per step).

Model parameters (quality / guidance / steps / seed, etc.)

Each model exposes its own inline setting pills in the bottom row (e.g. steps, guidance, resolution, seed). Click a pill for its control (slider, segmented buttons, or a Random/Fixed seed toggle), or scroll on the pill to adjust. A pill's value turns the category color when it differs from the model default; each has a per-pill reset.

Estimated cost

A coins count sits left of the Send button showing the estimated credit cost for the current configuration (model, image count, parameters).

Reset

Reset props (circular arrow, bottom row) — snaps model parameters, image count, and aspect ratio back to defaults. Only shows when something has changed.
Reset all (circular arrow, top-right) — also clears the prompt and removes all LoRAs. Board selection and the Design reference / Look / Matrix panels are left untouched.

Help mode

Open the … menu at the end of the bar and pick Explain the chat bar. While on, hovering any pill shows a description of what it does instead of opening its editor. Press Esc to exit. (This used to be a ? icon in the left icon rail; it moved into the menu with the rest of the bar's settings.)

Voice input

Dictation is the prompt field's left bookend — a microphone button inside the field, available whether or not the field has text in it. Recording is written down and fills the prompt: in Google Chrome by Chrome's own speech recognition, in other browsers on your device. Hold ⌘D (Ctrl+D) with the prompt focused to dictate without the button. Click the stop square to finish. (Send is a separate control and always shows an arrow.)

Collapse / expand & docking

Press Tab (when not typing in the prompt) to collapse or expand the bar, clearing it off the board.
The Position switch in the … menu moves the bar between bottom and left orientation. In the 2D editor the bar sits at the bottom, and the switch is shown disabled rather than hidden.

Editor2D editor: source switch ("Current View")

When editing an image in the 2D editor, the category pill is replaced by a mode pill:

Edit — generate from the visible canvas (image-to-image).
New — ignore the canvas and generate from the prompt alone (text-to-image).

Drawing a selection always takes priority, and the pill then offers two ways to use it:

Edit in selection — sends the whole picture; only what is inside the selection changes (inpainting).
Generate into selection — sends only what the selection shows, with your prompt and model (image-to-image), and the whole result comes back as a new layer filling the selection. Nothing is cut out of it. Most models draw only certain shapes, so the selection first snaps to the nearest one the chosen model draws (16:9, 4:3, 1:1 and so on), and you see it change; it grows around its centre so everything you selected stays inside.
Clear selection — goes back to Edit or New. Clearing the selection also puts the pill back on Edit in selection for the next one.

For a one-click sharper, more detailed version of a selection, right-click it and choose Upscale selection (see the 2D editor guide).