Frontier still direction
Generate stills with ChatGPT Image 2.5, Nano Banana 2, Nano Banana Pro, original Nano Banana, FLUX 2 Pro, Microsoft MAI Image 2.5 and 2.5 Pro, Seedream 5 Lite, Meta Muse Image, Grok Imagine Image 2.0, Qwen Image 3, and Kling Omni 3 from one workspace. Keep the image in your library.
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 endpointFrom the community
Open a real still, then Generate this in Image Lab.
A better stills surface
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.
Describe the still. Optionally attach a source image when you want ChatGPT Image 2.5, Nano Banana 2, Nano Banana Pro, MAI Image 2.5, MAI Image 2.5 Pro, Seedream 5 Lite, Muse Image, Grok Imagine Image 2.0, Qwen Image 3, or Kling Omni 3 to edit.
Choose ChatGPT Image 2.5, a Nano Banana SKU, FLUX 2 Pro, MAI Image 2.5, MAI Image 2.5 Pro, Seedream 5 Lite, Muse Image, Grok Imagine Image 2.0, Qwen Image 3, or Kling Omni 3 and an aspect ratio that model actually accepts.
When the still finishes, it lands in your Clip Studio Library.
The model slate
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 test03GoogleNano Banana 2Google Nano Banana 2 stills and edits on fal. Quoted at 1K.Text to imageImage to imageSee model details
Clip Studio test04GoogleNano Banana ProGoogle Nano Banana Pro stills and edits on fal. Quoted at 1K.Text to imageImage to imageSee model details
Clip Studio test05
Black Forest LabsFLUX.2 ProBlack Forest Labs production stills, billed by output megapixels.Text to imageSee model details
Clip Studio test06
MicrosoftMAI Image 2.5Microsoft photoreal stills via fal, including edits from a source image.Text to imageImage to imageSee model details
Clip Studio test07MicrosoftMAI Image 2.5 ProMicrosoft flagship stills via fal, including edits from a source image.Text to imageImage to imageSee model details
Clip Studio test08
ByteDanceSeedream 5 LiteByteDance Seedream 5 Lite stills and edits via fal. Aspect maps to a 2K-band size.Text to imageImage to imageSee model details
Clip Studio test09
MetaMuse ImageMeta Muse Image stills and instruction edits via fal. Around 2.5MP, with custom aspect strings.Text to imageImage to imageSee model details
Clip Studio test10
xAIGrok Imagine Image 2.0xAI Grok Imagine Image 2.0 stills and edits via fal on listed aspects.Text to imageImage to imageSee model details
Clip Studio test11
AlibabaQwen Image 3Alibaba Qwen Image 3 stills and edits via fal. Quoted at 1K, with strong typography and CJK.Text to imageImage to imageSee model details
Clip Studio test12
KuaishouKling Omni 3Kuaishou Kling Omni 3 stills and edits via fal. Quoted at 1K on listed Kling aspects.Text to imageImage to imageSee model detailsEditorial selection ยท Last reviewed 20 September 2026
Also in the studio
Before the first still
Image Lab is one workspace for generating stills across 12 image models in Clip Studio: OpenAI ChatGPT Image 2.5, Google Nano Banana, Nano Banana 2, Nano Banana Pro, Black Forest Labs FLUX 2 Pro, Microsoft MAI Image 2.5, Microsoft MAI Image 2.5 Pro, ByteDance Seedream 5 Lite, Meta Muse Image, xAI Grok Imagine Image 2.0, Alibaba Qwen Image 3, and Kuaishou Kling Omni 3. Write a prompt, pick a model and aspect ratio, generate, and keep the still in your library.
Ad Lab turns a website into locked ad layouts, plus cinematic tip packs, Proof 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.
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.
ChatGPT Image 2.5, Nano Banana 2, Nano Banana Pro, MAI Image 2.5, MAI Image 2.5 Pro, Seedream 5 Lite, Muse Image, Grok Imagine Image 2.0, Qwen Image 3, and Kling Omni 3 support image-to-image when a source still is attached. Original Nano Banana and FLUX 2 Pro are text-to-image here.
ChatGPT Image 2.5 (OpenAI), Nano Banana, Nano Banana 2, and Nano Banana Pro (Google), FLUX 2 Pro (Black Forest Labs), MAI Image 2.5 and MAI Image 2.5 Pro (Microsoft), Seedream 5 Lite (ByteDance), Muse Image (Meta), Grok Imagine Image 2.0 (xAI), Qwen Image 3 (Alibaba), and Kling Omni 3 (Kuaishou). Choose the model that fits your image.
Write a prompt, pick ChatGPT Image 2.5, Nano Banana, FLUX 2 Pro, MAI, Seedream 5 Lite, Muse Image, Grok Imagine Image 2.0, Qwen Image 3, or Kling Omni 3, and generate.
Enter Image Lab
