What's New

Flux Art Changelog tracks model launches, generation speed improvements, product workflow updates, and AI ecommerce creation features across Flux Art. Each entry records what changed, which models or workflows are affected, and how the update helps image generation, video generation, prompts, or ecommerce visual production.

Current

GPT Image 2.6 now supports animated GIFs and SVG vectors

Flare and Sunburst now offer more ways to create. Generate a looping GIF from a written idea, or make scalable SVG icons and illustrations alongside static images.

  • NewAnimated GIF | Describe the subject, movement and loop. Choose quality, aspect ratio or custom dimensions; current output uses 9 frames at 200ms per frame.
  • NewSVG vectors | Create icons, logo concepts and simple illustrations in SVG format, with quality, aspect ratio and custom dimensions available.
Create a GIFCreate an SVG

GPT Image 2.6 is here, with 50% off for a limited time

GPT Image 2.6 joins the workspace with Flare and Sunburst. Try a few directions with Flare, or refine an existing image with Sunburst. Both support image generation and editing up to 4K, with 50% off for a limited time.

  • New⚡ Flare | Turn ideas into images, then keep working on the direction you like. Great for everyday image creation, product images, and posters.
  • New☀️ Sunburst | Give the details a little more attention. Upload a reference image and use text to adjust backgrounds, materials, and details in the scene.
Try FlareTry Sunburst

Batch watermark: sign a whole set of images at once

Sign your work or give a set of product photos a consistent brand mark. Add text or a logo across the whole batch, with placement, size, and opacity set the way you like.

  • NewBatch watermark | Add a signature or logo to a set of images in one go. Choose a single watermark or a tiled layout, adjust its size and opacity, and process images for 5 credits each.
Add watermarks

Background removal: keep the subject you need

Take a person or product into your next design. Describe the subject to keep, and Background removal attempts to isolate it as a transparent PNG. Generative processing may change details, so compare the result with the original before use.

  • NewDescribe the person or product to keep. AI separates the subject from the background and exports a transparent PNG for posters, product images, and other designs.
  • ImproveImage generation now uses the same loading animation and transitions as the image workspace.
Try background removal

Object removal: clear away the distractions

A passerby or clutter on a table can draw attention away from the subject. Paint over the distraction, and Object removal will attempt to remove it and rebuild the background, bringing the focus back where you want it.

  • NewPaint over an unwanted object and let AI attempt to reconstruct the background. Instructions are optional; add a specific request such as “remove the flowers, keep the vase” when needed.
  • ImproveObject removal shares the selection editor with Local repaint, keeping brush, eraser, and zoom controls familiar.
Try object removal

Local repaint: mark the part you want to change

Sometimes one small edit makes the difference. Try a different chair, add flowers to an empty vase, or change a top. Mark the area, describe your idea, and keep refining the image with Local repaint.

  • NewPaint over the area to edit and describe what you want instead. Add a reference image to help AI understand the result you have in mind.
  • ImproveRefine selections with erase, undo, and zoom controls. Remove extra brush strokes or zoom in to work on smaller areas.
Try local repaint

Nano Banana 2.1 is here, with 50% off for a limited time

Give your next image a fresh starting point with Nano Banana 2.1. Create from text or bring a reference image to work from, then refine the background, style, and details. Output up to 4K is supported, with 50% off for a limited time.

  • New🍌 Nano Banana 2.1 | Turn ideas into images, or upload a reference and keep refining backgrounds, styles, and details. Supports output up to 4K, with 50% off for a limited time.
  • FixFixed several known issues to improve the overall experience.
Try Nano Banana 2.1Open Image Workspace

GPT Image 2.5 Flare and Sunburst are here, with 50% off for a limited time

Flare and Sunburst bring two more options to image creation. Use Flare to explore compositions and Sunburst to work through detailed edits. Both support generation and editing, with 50% off for a limited time.

  • New⚡ Flare | Quick everyday creation. Try different compositions, make product images or posters, then keep working on the direction you like.
  • New☀️ Sunburst | A closer focus on editing precision. Upload an image and describe what you want to change, from backgrounds and materials to details in the scene.
  • NewChoose from five quality levels, sizes up to 4K, and custom dimensions. Each model page has examples and prompts to help you find a starting point.
  • New🎁 Both models are 50% off for a limited time. Try Flare for quick image creation, or Sunburst for detailed edits.
Try FlareTry Sunburst

12 new AI E-commerce tools for product photos and virtual try-on

Product photos rarely stop at one image. Alongside Product Suite, 12 more tools now cover A+ Content, SKU variants, retouching, background changes, and virtual try-on. Choose the task you need from 13 e-commerce workflows.

  • NewCreate A+ Content and batch SKU images for product listings.
  • NewRecreate popular product visuals, retouch products, change colors, and replace backgrounds.
  • NewCreate apparel image sets, model try-on, accessory try-on, pose changes, face swaps, and shoe try-on results.
  • NewImproved: The existing Product Suite now supports more domestic and cross-border e-commerce platforms, with generated work easier to find in the E-commerce Asset Library.
  • FixFixed image upload and generation stability issues to improve the overall experience.
Create E-commerce ImagesView E-commerce Assets

Seedance 2.5 is here, with better AI E-commerce product suite quality

Seedance 2.5 is now available for creating AI videos up to 30 seconds from text, images, video, and audio, as well as editing or extending existing clips. AI E-commerce product suites now deliver better creation quality, while smoother page motion and a round of fixes make everyday creation more comfortable.

  • NewSeedance 2.5 is now available with multimodal references, video editing and extension, and AI video creation up to 30 seconds.
  • NewImproved: AI E-commerce product suite creation quality has been upgraded for better overall results.
  • NewImproved: Motion across most pages now feels smoother when browsing, opening dialogs, and choosing assets.
  • FixFixed multiple known issues affecting uploads, asset selection, and everyday creation workflows.
Try Seedance 2.5Try AI E-commerce

AI E-commerce product suites are now live

Build a product listing one set at a time. Upload a product photo, choose the platform and image types, and generate main images, white-background shots, selling-point images, and lifestyle scenes. Results stay together for review and export, with support for Taobao/Tmall, JD.com, Pinduoduo, and Douyin E-commerce.

  • NewCreate product listing images tailored to supported e-commerce platforms with AI E-commerce product suites.
  • NewReview and export a complete product suite, or start from a recommended example.
  • FixFixed known issues to improve page access and everyday creation workflows.
  • FixFixed browser compatibility issues.
Try AI E-commerce

19 Kling image and video models are now available in Flux Art

There is more room to explore with Kling: 8 image models and 11 video models now join Flux Art, covering V1–V3, Image O1, Video O1, and Omni workflows. Browse the model pages for a starting point, then head to the workspace to try your own idea.

  • NewEight Kling image models and eleven Kling video models are now available, giving creators more choices for image and video creation across a wider range of workflows
  • FixFixed multiple known issues to improve the overall experience
Explore Kling ModelsOpen Video Workspace

Asset Library is live for centralized creative asset management and reuse

Keep useful material close at hand. The Asset Library brings uploaded images, video, audio, and generated results into one private workspace, ready to find and reuse. Uploads, drag-and-drop, and media loading also received fixes in this update.

  • NewFlux Art Asset Library is live, bringing uploaded images, videos, audio, and AI-generated results into one place
  • FixFixed known issues with asset uploads, image drag-and-drop, and media loading to improve the overall experience
Open Asset Library

Fine editing is live for a better reference-image workflow

Some instructions are easier to draw than describe. Open a reference image to sketch, add text, or mark an area with arrows and shapes before generation. Fine editing is available from the homepage and both image and video workspaces, with undo and redo when you need another try.

  • NewFine editing is live with freehand drawing, text, arrows, multiple shape annotations, undo, and redo
  • NewReference images can be opened for fine editing from the image workspace, video workspace, and homepage creation entry
  • FixImproved reference-image source validation, loading, and upload handling, while fixing multiple known issues for better creation stability
Try Fine Editing

Flux Art OpenAPI is live, with an upgraded workspace task board

Bring generation into the tools you already use. Flux Art OpenAPI connects products, scripts, and automated workflows to models such as GPT Image 2, Nano Banana, and Seedance 2.0 through one server-side API. The workspace task board also has a refreshed layout for checking progress and results.

  • NewFlux Art OpenAPI is live, providing one API for multiple AI image and video generation models
  • NewOpenAPI overview, API Key management, and API reference pages provide a complete developer entry point
  • NewThe workspace task board now supports expanded and compact views, making multi-task creation and management clearer and more efficient
  • FixFixed multiple known issues to improve page access, account sessions, and creation workflow stability
View OpenAPIOpen Image Workspace

More model topic pages and smoother drag-and-drop uploads

More model pages now include a place to start, and reference files are easier to bring along. Drop files directly into the image or video workspace, or browse the new Midjourney, Wan, and Seedance topic pages before choosing a model.

  • NewTopic pages are now live for Midjourney V7 Blend, Wan 2.2 Plus, Seedance 1.0 Pro, Seedance 1.0 Pro Fast, Seedance 1.5 Pro, and Seedance 2.0 Fast
  • NewAI image and video workspaces now support drag-and-drop uploads for easier asset input
  • FixFixed video playback, asset links, and layout issues on the Seedance 2.0 Mini topic page, with additional interaction refinements
View Models

Seedream 5.0 Pro and Seedance 2.0 Mini are live, upgrading image and video creation

This update brings Seedream 5.0 Pro and Seedance 2.0 Mini to Flux Art, together with complete topic content for both models. Seedance 2.0 now supports 1080p and 4K output, while Seedream 5.0 gains a more flexible multi-resolution generation experience.

  • NewSeedream 5.0 Pro is live with 1K / 2K AI image generation and editing for infographics, precise revisions, commercial visuals, and multilingual design
  • NewSeedance 2.0 Mini is live with text-to-video, image-to-video, multimodal references, video editing and extension, optional audio, and 4–15 second output up to 720p
  • NewDedicated topic content is live for Seedream 5.0 Pro and Seedance 2.0 Mini. Seedance 2.0 now supports 1080p and 4K resolutions for higher-quality AI video generation
  • ImproveOptimized image loading and download strategies to improve loading speed and enhance the overall browsing experience
  • ImproveImproved Seedream 5.0 generation across 1K, 3K, and 4K output sizes
View Seedream 5.0 ProView Seedance 2.0 Mini

HappyHorse 1.0 / 1.1 video models are live, upgrading AI video creation

This update brings HappyHorse 1.0 and HappyHorse 1.1 to Flux Art, expanding AI video generation with text-to-video, image-to-video, and reference-driven motion workflows for product clips, brand visuals, social videos, and creative storyboards. We also added topic content for Wan 2.2 Flash and Wan 2.5 Image-to-Image, while polishing several details across the creation flow.

  • NewHappyHorse 1.0 / 1.1 video models are live, expanding AI video generation and reference-based creation
  • NewHappyHorse topic content is live for product clips, brand visuals, audio-aware videos, and motion storyboards
  • NewWan 2.2 Flash and Wan 2.5 Image-to-Image topic content added for fast drafts and reference-image editing
  • FixFixed multiple known issues to improve model selection, asset management, and creation stability
Try HappyHorse 1.1

Nano Banana 2 Lite is live, with new Grok topic pages

This update brings Nano Banana 2 Lite to Flux Art for fast 1K drafts, lightweight image editing, ecommerce hero image concepts, and social visual directions. We also added topic content for Grok Imagine and Grok Imagine Image Pro, expanding AI image generation scenarios for product visuals, campaign posters, and reference-based edits. Platform stability was improved as well, with better error reporting, network retries, and upload reliability.

  • NewNano Banana 2 Lite is live for fast 1K image drafts and lightweight editing
  • NewGrok Imagine and Grok Imagine Image Pro topic pages added for product visuals, posters, and reference-based edits
  • ImproveImproved error reporting, network retries, and OSS upload reliability for AI image generation workflows
  • FixFixed multiple known issues to improve browsing, model usage, and creation stability
Try Nano Banana 2 LiteView Grok ImagineView Grok Imagine Image Pro

GPT Image 2 custom sizes are here, with faster generation

This update upgrades Flux Art's AI image generation workflow. GPT Image 2 now supports custom sizes for e-commerce hero images, brand posters, social visuals, and detail-page assets, while interface responsiveness and image generation speed have both been improved for a smoother creative flow.

  • NewGPT Image 2 now supports custom sizes for more flexible output specs
  • ImproveReduced interface stutter across browsing and creation flows
  • ImproveImproved AI image generation speed to shorten waiting time
  • FixFixed multiple known issues to improve overall stability
Try GPT Image 2

New Inspiration page and workspace, smoother AI creation

This update upgrades the Flux Art experience across AI image generation, AI video generation, and prompt-based creation. The new Inspiration page adds more curated prompt templates for e-commerce hero images, brand posters, social visuals, and short-video storyboards, while the new workspace makes GPT Image 2, Nano Banana, Seedance 2.0, and other models easier to use in one flow.

  • NewInspiration page launched with a larger prompt library for faster creative direction
  • NewWorkspace launched with clearer image generation and video generation flows
  • NewExperience designs and topic pages updated for more AI creation scenarios
  • FixFixed known issues to improve page interactions and generation stability
View Models

Faster site-wide, more stable AI image & video generation, plus Feedback and Invoice requests

This release is all about a smoother ride. We ran a site-wide performance pass on assets and APIs so pages load faster, and tightened the reliability of both AI image generation and AI video generation. Whether you're a designer building e-commerce hero images, a cross-border team generating in bulk, or a motion-comic or short-video creator on a deadline, every step from picking a model to exporting feels more responsive. We also shipped in-app Feedback and Invoice Request, so requests and reimbursements stay in one place.

  • ImproveSite-wide load speed; stability and interaction polish across AI image and AI video generation
  • NewFeedback entry point, Invoice Request entry point
Open AI ImageTry AI Video

Concurrency scaled up, GPT Image 2 image generation is now faster

As Flux Art usage climbed, GPT Image 2 quickly became one of the most-used models on the platform. We expanded concurrency capacity and pushed up GPT Image 2 generation speed, so high-volume jobs like e-commerce detail images, posters, and social posts spend less time in the queue and come out more consistently.

  • ImproveExpanded concurrency capacity, improved GPT Image 2 (gpt-image-2) generation speed
Try GPT Image 2

Nano Banana 2, Nano Banana Pro, and Nano Banana sped up across the board

The Nano Banana series has long been a go-to for its strong price-to-quality ratio and dependable output. This update speeds up the entire Nano Banana lineup - Nano Banana 2, Nano Banana Pro, and Nano Banana all respond faster, keeping batch work for e-commerce, design, and social content moving.

  • ImproveImproved generation speed for Nano Banana 2, Nano Banana Pro, and Nano Banana

GPT Image 2 Prompt Library is live - get great prompts with zero experience

A lot of new users get stuck on what to actually type. To lower the bar for AI image generation, Flux Art added a GPT Image 2 Prompt Library to the Inspiration page, packed with ready-to-use GPT Image 2 prompt templates. Product hero shots, posters, portraits, scene images - designers and e-commerce sellers can grab a prompt, tweak the details, and ship.

  • NewInspiration page now includes a large set of GPT Image 2 prompts spanning e-commerce, design, and portrait use cases

Creative Templates launched - one-click starts for e-commerce, cross-border, social, and design

Writing prompts from scratch is slow; templates are where the speed is. Flux Art launched Creative Templates that turn high-frequency jobs - e-commerce hero images, detail shots, promo posters, social media graphics - into one-click starting points. E-commerce sellers, cross-border merchants, social media managers, and designers can get going with zero ramp-up: pick a template, swap the content, export.

  • NewCreative Templates covering e-commerce, cross-border, social media, and design workflows

Grok Imagine series added (Grok Imagine Image / Grok Imagine Image Pro)

More models, more room to create. Flux Art added the latest Grok Imagine series - Grok Imagine Image and Grok Imagine Image Pro - bringing richer stylization and creative range, and giving designers and content creators another set of options that just works.

  • NewGrok Imagine Image and Grok Imagine Image Pro added to the model lineup

Seedream 5.0 added - web-connected generation for higher-quality e-commerce output

Flux Art added Seedream 5.0 with support for web-connected AI image generation, pulling more accurate reference into the frame to meet higher-quality needs for e-commerce hero images, detail shots, and brand visuals. For teams and designers chasing output quality, it's a strong new option.

  • NewSeedream 5.0 added, with web-connected generation aimed at high-quality e-commerce output

Wan series and Qwen series added - image generation and image editing

Two widely used model families landed at once. Flux Art added the Wan series and Qwen series, covering both AI image generation and AI image editing, so e-commerce, design, and social creators have the right model for each task - generate, retouch, and do localized edits, all on one platform.

  • NewWan series image generation and editing models added
  • NewQwen series image generation and editing models added

GPT Image 2 and the full Nano Banana series, 50% off (limited time)

To thank our early users, Flux Art ran a limited-time 50% off on two of the most popular models: GPT Image 2 and the full Nano Banana series (including Nano Banana 2 and Nano Banana Pro). Whether you're an e-commerce team batching out hero images or a designer and social creator generating daily, the same budget went a lot further.

  • New50% off on GPT Image 2 and the full Nano Banana series, limited time

Midjourney V7 series added

Flux Art added the Midjourney V7 series, bringing its signature artistic quality and composition into the platform. No separate subscription or setup required - creators can call Midjourney V7 directly inside Flux Art for posters, concept art, and brand visuals, ready to use.

  • NewMidjourney V7 series models added

AI Video workspace is live - Seedance 2.0 lineup plus Grok Video

From "make images" to "make video." Flux Art launched its AI Video workspace, supporting Seedance 2.0, Seedance 1.5, Seedance 1.0, and Grok Video in one place. E-commerce teams can produce hero-image videos and short-form assets, motion-comic and webtoon creators can turn storyboards into shots, and social creators can spin up clips fast.

  • NewAI Video workspace launched, supporting Seedance 2.0 / 1.5 / 1.0 and Grok Video
Try AI Video

Flux Art launches - debuting GPT Image 2 and the full Nano Banana series

Flux Art is live: one platform aggregating 50+ leading AI image and video models, launching with two of the most in-demand - GPT Image 2 and the full Nano Banana series (Nano Banana 2, Nano Banana Pro, Nano Banana). Roughly 20-second generation, 4K watermark-free output cleared for commercial use, free credits on sign-up, every major model in a single account with no separate subscriptions, and your data is never used for training. E-commerce sellers, designers, cross-border teams, and content creators can start generating on day one.

  • NewGPT Image 2 and the full Nano Banana series launched, kicking off AI image generation on Flux Art
Start Creating

Flux Art's next horizon goes far beyond this...