01
Write a prompt
Describe the still. Optionally attach a source image when you want ChatGPT Image 2.5, MAI Image 2.5, or MAI Image 2.5 Pro to edit.
Frontier still direction
One prompt.Five image instincts.
Generate stills with ChatGPT Image 2.5, Google Nano Banana, FLUX 2 Pro, and Microsoft MAI Image 2.5 and 2.5 Pro from one workspace. Keep the image in your library.
Quote before you generateChatGPT Image 2.5 and MAI can edit a stillResults saved to your library
Image Lab / Model stage
5 onlineText to imageImage to imageProduct shots and image-to-image edits.
Director note
OpenAI ChatGPT Image 2.5 stills via fal, including edits from a source image.
Clip Studio test. Results vary with prompt, settings, and provider updates.
fal endpointModesText to image · Image to imageFrame9:16 · 1:1 · 4:5 · 16:9Best forProduct shots and image-to-image edits.OutputStill
From the community
Published Image Lab stills.
Open a real still, then Generate this in Image Lab.
A better stills surface
Model choice without playground chaos.
Image Lab is still generation: you leave with a downloadable image, not an ad set or a motion clip. Prompt, model, aspect, and quote stay in one room.
02
Pick a model and aspect
Choose ChatGPT Image 2.5, Nano Banana, FLUX 2 Pro, MAI Image 2.5, or MAI Image 2.5 Pro and an aspect ratio that model actually accepts.
03
Generate and keep the still
When the still finishes, it lands in your Clip Studio Library.
The model slate
5 distinct still instincts.
Clip Studio test01
OpenAIChatGPT Image 2.5OpenAI ChatGPT Image 2.5 stills via fal, including edits from a source image.Text to imageImage to imageSee model details
Clip Studio test02
GoogleNano BananaGoogle Nano Banana stills on fal. Flat per-image price.Text to imageSee model details
Clip Studio test03
Black Forest LabsFLUX.2 ProBlack Forest Labs production stills, billed by output megapixels.Text to imageSee model details
Clip Studio test04
MicrosoftMAI Image 2.5Microsoft photoreal stills via fal, including edits from a source image.Text to imageImage to imageSee model details
Clip Studio test05MicrosoftMAI Image 2.5 ProMicrosoft flagship stills via fal, including edits from a source image.Text to imageImage to imageSee model detailsEditorial selection · Last reviewed 9 September 2026
Also in the studio
Image Lab sits next to motion and voice.
Before the first still
Questions from the stills room.
What is Clip Studio Image Lab?
Image Lab is one workspace for generating stills across 5 image models in Clip Studio: OpenAI ChatGPT Image 2.5, Google Nano Banana, Black Forest Labs FLUX 2 Pro, Microsoft MAI Image 2.5, and Microsoft MAI Image 2.5 Pro. Write a prompt, pick a model and aspect ratio, generate, and keep the still in your library.
How is Image Lab different from Ad Lab or AI Stills Video?
Ad Lab turns a website into locked ad layouts, plus cinematic tip packs and Explain films. AI Stills Video turns a generated frame into a motion clip. Image Lab is still generation: you leave with a downloadable image, not an ad set or a film.
How is Image Lab priced?
Each generation is quoted before you run it. Model, aspect ratio, and whether you attach a source still change the quote. Each generation uses your monthly generation allowance.
Can I edit an existing still?
ChatGPT Image 2.5, MAI Image 2.5, and MAI Image 2.5 Pro support image-to-image when a source still is attached. Nano Banana and FLUX 2 Pro are text-to-image here.
Which models are included?
ChatGPT Image 2.5 (OpenAI), Nano Banana (Google), FLUX 2 Pro (Black Forest Labs), MAI Image 2.5 (Microsoft), and MAI Image 2.5 Pro (Microsoft). Choose the model that fits your image.
Bring the still.Choose the instinct.
Write a prompt, pick ChatGPT Image 2.5, Nano Banana, FLUX 2 Pro, MAI Image 2.5, or MAI Image 2.5 Pro, and generate.
Enter Image Lab