AI Box

Generation language

Default language for AI-written scripts, TTS voiceovers, and captions. You can still override per-job.

multi-image editor

Edit images by chatting.

The Image Agent is a conversational image editor. Drop an image, type what to change ('put her in a yellow jacket', 'remove the background guy'), and the right model handles it. We pick FLUX Kontext for surgical edits, Nano Banana for multi-image composition, Seedream for photoreal, Qwen for cheap iteration.

cost

Model-weighted (2-4 credits)

powered by

FLUX Kontext Pro / Max · Nano Banana (Gemini 2.5) +5

still from workspace

why it matters

Built for the way creators actually work.

benefit 01

8 edit models, auto-routed

Multi-image input? Routes to Nano Banana. Typography-heavy edit? Ideogram. Cheap iteration? Qwen. You can override the pick, but the default is usually right.

benefit 02

Chat history

Every iteration is a turn. Edit, see the result, refine. Each turn is its own image you can save, download, or branch from.

benefit 03

Save to Library

Send any edit to your Asset Library or My Brand → Assets for reuse across renders. Sample images are 4 generated images you can hand off to Multi-Scene UGC as B-roll.

benefit 04

Brand integration

When My Brand has a product catalog, you can mention products by name and the agent finds the reference image automatically.

how it works

4 steps.
Most of it is the model, not you.

Drop image + describe the change

Single image edit ('add a logo to the cup'), multi-image composite ('put this person in this scene'), or generation from scratch ('product on a marble countertop, golden hour').

Agent picks the right model

FLUX Kontext for surgical edits, Nano Banana for multi-image, Seedream for photoreal, Recraft for vector / brand-safe — auto-routed based on your input.

Iterate via chat

Each turn is a new edit on top of the last result. Branch any turn into a separate thread, or roll back if you don't like the direction.

Save or send

Save to your Asset Library, send to My Brand → Products, or use as B-roll in the next Multi-Scene UGC render.

built for

Made for the people shipping shorts.

Product photographers (without a studio)

Take one product shot, render 20 lifestyle variants without ever setting up a backdrop.

Hero-shot designers

Generate the hero, then refine the wardrobe, lighting, and pose iteratively. The chat keeps the subject consistent.

Marketing teams

Composite a person into a scene, swap backgrounds, change brand colors — all by typing.

Creators with brand assets

Apply your brand colors and logo to generated images via the prompt. My Brand integration finds your kit automatically.

models in the loop

FLUX Kontext Pro / MaxNano Banana (Gemini 2.5)Seedream 4Qwen Image EditIdeogram V3Recraft V3FLUX Dev img2img

briefing notes

Questions worth answering.

play in the same band

Pairs well with.

no card · cancel anytime

Open Image Agent in 60 seconds.

Free plan includes 50 credits and the workspace, projects, brand kit, and saved prompts. Upgrade when you outgrow it.

See pricing