
Show HN: A web workspace for trying FLUX 3 Klein
Turn your ideas into images with FLUX 3 Klein.
以下正文同步自 Hacker News 最新,版权归原站所有,已转换为易读排版。
A LITTLE PROMPT. A BIG IDEA.
Turn your ideas into images with FLUX 3 Klein.
Create beautiful images for your social posts, product showcases, blog covers, and personal projects. Give your products a polished look, help readers picture your story, and turn your ideas into something worth sharing.
WORDS TO IMAGES EDIT & COMBINE NATIVE UP TO 4K
Create an imageStart with your wordsEdit & combineStart with your images
What are you imagining?0/4000
Try:Studio portraitProduct photoStorybook artTypography poster
Aspect ratio
Klein is trained on detailed scene captions. Keep enhancement on for short prompts; use original wording when you have already written a complete scene description.
Reuse a seed with the same prompt and settings for more consistent variations. Leave blank to explore.
Loading inputs…We protect your privacy; only successful images use credits.
PhotographyClay illustrationLayout design
Your generated image will appear here.
Describe the subject, setting and lighting you want.
When editing, say what to change and what to keep.
Give each reference a role: “Image 1 for the person, Image 2 for the background.”
Browse examples & prompts
Find your next idea
Photography, illustration and graphic design. Open an image, discover its prompt, and make it your own.
TextureNeon letteringCollageAnime atmosphereCharcoal drawingCharacter portraitVisual instructionsLayout designFashion editorialPortraitGolden hour
← Previous1 / 3Next →
Precise image editing
Target individual parts or request several changes together while retaining the surrounding scene. Drag each comparison to see the original and edited example.
Give this tree-lined street a light winter snowfall. Preserve the vehicles and street layout.
Turn the waterfall into ice and long icicles while keeping its surrounding forest.
Add a small red bandana around the dog’s neck. Keep the dog and the rest of the scene.
Move the blue convertible to an empty desert road beneath a clear sky.
Place three camels along the distant dune ridge while retaining the sunset landscape.
Add wide crimson stripes to the sails while preserving the boat and sea.
Generation, editing and multiple references share one model. This workspace accepts up to four reference images.
Choose 1K, 2K or 4K output directly from the model. The selected aspect ratio sets your canvas.
Klein expects detailed scene captions. Prompt enhancement expands short instructions, incorporating reference images when editing. It is enabled by default here.
Bounding boxes let you specify where objects and text belong, or target individual regions for edits.
Region drawing is not available in this workspace yet. Use text instructions here, or see the official layout documentation.
Start with a prompt, your own photo or an example from the gallery. Create social images, product concepts and blog illustrations in three steps.
Choose Create an image and describe your subject, composition, lighting and style. For FLUX 3 Klein image editing, choose Edit & combine and upload up to four references. You can also open a gallery example and select Create similar to load its image and prompt.
Pick a square, portrait or landscape canvas, then choose 1K, 2K or 4K. Start at 1K to explore an idea. Select PNG, JPEG or WebP for your download, and keep prompt enhancement on if you want help expanding a short description.
FLUX 3 Klein vs Nano Banana 2.1 and FLUX 2 Klein. Three models. Different ways to create.
For creating and refining images from a prompt and your references.
Consider it for search-grounded imagery and larger sets of references.
Consider it for compact generation and editing at smaller output sizes.
Documented capabilities checked October 17, 2026. This is a feature comparison, not a head-to-head quality benchmark. Reference limits depend on the model and provider; this workspace accepts four images. 4K describes image dimensions; 4MP describes total pixel count.
FLUX 3 Klein brings 5.6B parameters to a consumer GPU. Choose distilled weights for faster inference, or Base weights for fine-tuning and LoRA training.
Build with the API, Hugging Face or ComfyUI under the applicable license. Read the official technical guide ↗
Your photos and prompts are shared with our AI providers for safety checks and image generation. Uploads and results are stored until you request deletion. We won't use them in our gallery or advertising without your separate permission. Anyone with a media link may be able to view it. To request deletion, email contact@flux3klein.net. Read our Privacy Policy for details.
Yes. You can use a generated image as a reference for another edit, such as changing the background, colors or details. Open your result, choose Edit image, and describe what you want to change. Keep refining it with new instructions, then download the version you want. Loading an image for editing is free; each new generation uses credits only when it succeeds.
FLUX 3 Image Klein is an image model from Black Forest Labs. It supports text-to-image generation and image editing. This independent website gives you a browser workspace for the model through fal.ai.
Choose Text to image, describe your idea, select an aspect ratio and resolution, then sign in to generate. To edit an existing image, switch to Edit image, add up to four references, and describe the finished result you want.
Yes. Choose Edit image and upload up to four JPG, PNG or WebP images, each under 30 MB. Describe both the change and the elements you want to keep. You can combine references by describing how each selected image should contribute.
Choose 1K, 2K or 4K. These correspond to a longest edge of 1024, 2048 or 4096 pixels. The other dimension follows your selected aspect ratio. Each request generates one image in your selected PNG, JPEG or WebP format.
Yes. New users can generate 5 free 1K images with the 10 welcome credits granted once on sign-up. No credit card is required. After that, buy a one-time credit pack. A 1K image uses 2 credits, 2K uses 5, and 4K uses 10. Signing in again does not grant another welcome gift.
Describe the subject first, then add the setting, composition, lighting and visual style. For example: a ceramic coffee cup on a pale oak table, close-up composition, soft morning window light, editorial product photography. For edits, say what should change and what should stay the same. Put any text you want in the image in quotation marks and check the spelling in your result.
FLUX 3 Klein supports native 4K generation and API prompt enhancement. FLUX 2 Klein supports outputs up to 4 megapixels and does not include built-in prompt upsampling. Both support text-to-image and reference-based editing. This FLUX 3 Klein workspace accepts up to four references.
Choose based on your workflow. This site offers FLUX 3 Klein generation and editing with up to four references and 1K, 2K or 4K output. Nano Banana 2.1 supports Google Search grounding, conversational editing and up to 14 reference images through Google's tools. Try the same brief in each model to compare the results you prefer.
Credits are deducted only when a generation completes successfully. Failed or blocked requests do not use your credits. You can find completed results in My Creations.
Check the applicable model and provider terms for your intended use. You are responsible for having the necessary rights to uploaded images, brands, people and other content. AI output can contain errors and is not guaranteed to be unique or free of third-party rights.
No. flux3klein.net is independently operated by Mizhi Smart. We are not affiliated with, sponsored by, or endorsed by Black Forest Labs. For official model documentation, weights and licensing, visit bfl.ai.
正文由 探How 从来源站点 RSS 同步,内容未经改写;遇到排版缺失或需要图片、视频时请以原文为准。