Updated weekly

Moduvo release notes

What's new in Moduvo: the latest AI models, features and improvements, every week. For what is coming next, see the roadmap.

See the roadmap
  1. Image Generator

    Product Photos with Nano Banana 2.1 and GPT Image 2.5

    Product Photos now runs on Nano Banana 2.1 and GPT Image 2.5, so your product shots get cleaner detail and sharper text. You can also pick several seasonal themes at once and get one photo for each. If an image takes longer than usual, you now see a clear message instead of a stuck progress bar.

    • New

      Newest models in Product Photos

      Choose Nano Banana 2.1 or GPT Image 2.5 in Listing Pack, Batch and Dimension Transformer. Nano Banana 2.1 offers 1K, 2K and 4K. GPT Image 2.5 offers low, medium and high quality at 1K or 2K. The credit estimate updates as you change the settings.

    • New

      Several themes in one go

      Seasonal Theme now lets you select more than one theme and creates one photo per theme, one after another. A new Easter theme is included, and the summary shows how many themes finished.

    • Improved

      Compare shows the live models

      The side-by-side comparison now lists the image models you can use today, including GPT Image 2.5, Nano Banana 2.1, Ideogram 4.0 and Seedream 5.0 Pro. GPT Image 2.5 offers 8 image shapes.

    • Improved

      Clear message when an image is slow

      After 3 minutes, a running image shows that it is taking longer and will appear in your gallery when ready. You can hide it, or cancel it and get the reserved credits back. A model that is temporarily down is now named as unavailable.

    • Improved

      Video to Video reads your clip

      Aleph 2.0, Kling 4.0 Edit and Seedance 2.5 Edit now show the detected shape of your source video, for example 9:16 (vertical). Phone clips keep their upright orientation. Large files and frame rates are handled for you.

  2. Animation Creator

    Kling 4.0: 30-second videos and video edit

    Kling 4.0 is now in Animation Creator with four ways to work: text, image, reference and edit. Clips run 3 to 30 seconds in 720p or 1080p, and you can steer them with up to 10 images or 5 videos. The Video to Video tab now lets you pick Aleph 2.0, Kling 4.0 Edit or Seedance 2.5 Edit and keeps your video and prompt when you switch.

    Good to know

    Kling AI, the video unit of Kuaishou, launched Kling 4.0 in October 2026, according to press coverage. It is known for clips of up to 30 seconds guided by as many as 10 keyframe images. Tip: for a product ad, add your packshot as Image 1 and write "Image 1 rotates slowly on a marble table".

    • New

      Kling 4.0 in four modes

      Make a video from text, from a first frame, from references or by editing a clip. Choose 3 to 30 seconds, 720p or 1080p, and 16:9, 9:16, 1:1 or 21:9 for text and reference videos.

    • New

      Up to 10 frames in one clip

      Set a first frame, a last frame and up to 8 middle frames. They are spaced evenly through the clip, so you decide how the story moves.

    • New

      Reference with images and videos

      Add up to 10 images and 5 videos (30 seconds in total). Click a tag such as Image 1 or Video 2 to place it in your prompt, and hover it to see the file.

    • New

      Edit a video with Kling 4.0

      Upload a 3 to 30 second MP4 or MOV, or pick one from History, and describe what should change. You can add up to 10 style images, and the result keeps its original sound.

    • Improved

      One place to edit videos

      The Video to Video tab now offers Aleph 2.0, Kling 4.0 Edit and Seedance 2.5 Edit. Your video and prompt stay when you switch, and the page tells you if a clip is too long or large for a model.

    • Improved

      Previews on prompt tags

      Tags in your prompt now show a small thumbnail of the photo or video they point to, so you always know which file is which.

  3. Image GeneratorAnimation Creator

    Edit any image, extend Omni videos

    Image Generator now has a full Edit page, so you can keep refining a picture version by version, and it works with GPT Image 2.5 and several other models. In Animation Creator you can edit and extend your Gemini Omni videos, and the best-ranked models are now marked on their cards.

    Good to know

    Google released Gemini Omni 1.1 Flash on August 27, 2026, with scene extension and first-and-last-frame control. In public blind tests on Arena, OpenAI's GPT Image 2.5 Sunburst ranked first for text-to-image in the October 7 leaderboard (preliminary at the time of writing). Tip: for an edit, keep the instruction short and add "keep everything else the same".

    • New

      Edit page for images

      Click Edit under a finished image to open a full workspace. Add extra photos, describe the change and see every version in a strip. Works with Nano Banana 2.1, GPT Image 2.5, GPT Image 2, Gemini 3 Pro, Nano Banana 2, Grok and Seedream 5 Pro.

    • New

      Download every version

      Save any version as PNG, JPG or WebP, or download all versions in one ZIP.

    • New

      Extend Omni videos

      Continue a Gemini Omni video by 4, 6, 8 or 10 seconds at a time, up to 40 seconds in total. The new Edit / Extend screen also lets you edit videos up to 10 seconds with up to 6 reference images, and add camera moves like push in, orbit or handheld with one click.

    • New

      More control over GPT Image 2.5

      Choose PNG, JPG or WebP output, switch on a transparent background for logos and stickers, set compression, and pick the new Extra high quality level at 1K, 2K or 4K.

    • New

      Best-ranked models are marked

      Model cards for GPT Image 2.5, Nano Banana 2.1, Gemini Omni and Seedance 2.5 show a Best quality badge. Other top models, such as Wan 3.0, show Top pick. The badges are based on public blind-test rankings from October 2026.

    • Improved

      Veo 3 has been retired

      Veo 3 is no longer available because its provider retired it. Your existing Veo 3 videos stay in History and can still be downloaded.

  4. Image Generator

    Nano Banana 2.1 in Image Generator

    Google's newest fast image model is now in Moduvo. It is sharper, follows prompts more closely and keeps faces and products consistent across a whole series, which makes it a great default for campaign visuals, banners and quick edits.

    Good to know

    Google released Nano Banana 2.1 on October 6, 2026 as the upgrade to Nano Banana 2. It improves image quality, text rendering and character consistency, and fixes the artifacts older versions produced in very wide 2K and 4K formats.

    • New

      Up to 4K in 14 formats

      Generate in 1K, 2K or 4K, from square posts and 9:16 stories to panoramic 8:1 website banners and tall 1:8 skyscrapers.

    • New

      Up to 14 reference images

      Add your products, faces, logos or a style reference to one generation and keep a whole campaign on brand.

    • New

      Search grounding and thinking mode

      Let the model look up current information before it draws, or turn on thinking for complex scenes and text-heavy layouts.

    • New

      Edit by chatting

      Ask for changes to the image you just made. Moduvo keeps the context, so you refine the result instead of starting over.

    • Improved

      Fairer credits

      You only pay for what the image actually used. If a generation stalls, it is cancelled automatically and the credits return to your workspace.

  5. Animation CreatorImage Generator

    Wan 3.0 video and GPT Image 2.5

    Two big model updates in one release. Wan 3.0 creates clips up to 30 seconds and can build a video from your own documents and media, and OpenAI's GPT Image 2.5 brings faster, more precise image editing.

    Good to know

    Alibaba launched Wan 3.0 on August 24, 2026. It doubles the 15-second limit of Wan 2.7 and keeps characters and scenes consistent across the whole clip. OpenAI released GPT Image 2.5 on September 8, 2026 in two versions: Flare for speed and Sunburst for detailed, precise edits.

    • New

      Wan 3.0 and Wan 3.0 Prime

      Create 2 to 30 second clips from text or an image in two quality tiers, with optional audio. Auto mode picks the right length for your prompt.

    • New

      Turn your materials into video

      Guide a Wan 3.0 clip with up to 10 images, 5 video clips and 5 audio clips, or attach a brief as a document or a link.

    • New

      GPT Image 2.5: Flare and Sunburst

      Pick Flare for results in seconds or Sunburst for precise edits, in 8 aspect ratios, 1K to 4K and four quality levels.

    • Improved

      Clear limits before you generate

      If an attachment is too large or too long, Moduvo tells you which file and which limit before any credits are used.

    • Improved

      Progress you can trust

      Video previews reconnect when you come back to the tab, and long renders show a clear status instead of looking stuck.

  6. Animation Creator

    Edit and extend your videos with Seedance 2.5

    You no longer have to regenerate a whole clip to change one thing. Describe the change and Seedance 2.5 edits the existing video, or continues it where it ends to build longer scenes.

    • New

      Edit video

      Pick any clip from your history and describe what should change. The rest of the shot stays as it was.

    • New

      Extend video

      Continue a clip from its last frame. The original and the new footage are joined into one finished video.

    • New

      Auto duration

      Not sure how long the clip should be? Auto mode chooses the length that fits your prompt.

    • Improved

      Faster history

      Animation history now loads in pages, so even large libraries open quickly.

  7. Animation Creator

    Seedance 2.5: 30-second videos in one take

    Seedance 2.5 is now in Animation Creator. It creates up to 30 seconds of video with sound in a single generation, so ads, product demos and short stories no longer have to be stitched together from short clips.

    Good to know

    Seedance 2.5 was released at the end of July 2026 as the next version of the Seedance video model. Its headline feature is a 30-second clip in a single pass, guided by many reference images, videos and audio clips at once.

    • New

      4 to 30 seconds with native audio

      Text-to-video and image-to-video in 480p, 720p or 1080p, with sound generated together with the picture.

    • New

      First and last frame

      Start from your own image and optionally set the final frame, so the clip lands exactly where you want it.

    • New

      Seedance 2.5 Reference

      Guide one video with up to 30 images, 10 videos and 10 audio clips, for consistent characters, products and voices.

    • New

      Up to 8 shots in one clip

      Scene Director lets you plan up to 8 shots within 30 seconds, each with its own description.

    • New

      Web search grounding

      Let the model look up current information for text-to-video prompts about real products, places or events.

  8. Image Generator

    More reliable image editing

    Editing images with Gemini 3 Pro and Nano Banana 2 is now much more reliable, especially with large photos as references.

    • Improved

      Large photos just work

      Big reference images are resized automatically in your browser before upload, so phone photos and high-res product shots no longer break a generation.

    • Improved

      Latest Gemini image model first

      Gemini 3 Pro edits now run on Google's stable release, with an automatic fallback if it is busy.

    • Fixed

      Clear reasons instead of empty results

      If a request is blocked, you see why, instead of a generic "no images" message.

    • Fixed

      No more endless spinners

      A generation that gets stuck is stopped automatically after 15 minutes instead of spinning forever.

  9. Image GeneratorSmart Rewrite Tool

    Seedream 5.0 Pro and a model picker in Smart Rewrite

    A new flagship image model for clean text and product detail, and the freedom to choose which AI writes for you in Smart Rewrite.

    Good to know

    Seedream 5.0 Pro is the flagship of the Seedream image family. It combines generation and reference-based editing in one model and is known for legible small text and logos, which makes it strong for packaging, ads and product shots.

    • New

      Seedream 5.0 Pro

      Create or edit images in 1K or 2K across 11 aspect ratios, including cinematic 21:9 and 9:21.

    • New

      Up to 10 reference images

      Combine products, people and styles from several images into one consistent result.

    • New

      Choose your writing model

      Smart Rewrite now lets you pick the provider and the exact model: OpenAI GPT-5.6, Claude Haiku, Sonnet or Opus, Gemini, Grok or Perplexity Sonar for answers with web sources.

  10. Animation Creator

    Wan 2.7 and Gemini Omni Flash for video

    Two new video models in Animation Creator. Wan 2.7 is great for multi-character scenes up to 15 seconds, and Gemini Omni Flash lets you change a finished video just by describing what should be different.

    Good to know

    Gemini Omni Flash is the first model of Google's new Omni family, announced at Google I/O 2026. It combines video generation and conversational editing in one model. Wan 2.7 from Alibaba brought reference-to-video with up to five consistent characters.

    • New

      Wan 2.7

      Text-to-video and image-to-video up to 15 seconds with optional audio. In image-to-video you can also set the last frame.

    • New

      Wan 2.7 Reference

      Use up to 5 images or videos of your characters and settings to keep them consistent in one scene. The starting frame is optional.

    • New

      Gemini Omni Flash

      Create 4 to 10 second videos from text or an image in 16:9 or 9:16, guided by up to 6 references.

    • New

      Edit generation

      Open a Gemini Omni Flash video from history, describe what to change and get a new version that continues from the original.

    • Improved

      Clearer content messages

      If a prompt is rejected by the model's content rules, you now see an explanation instead of a silent failure.

  11. Animation CreatorImage GeneratorVoice Generator

    Asset Library, Pro Mode and smarter prompts

    A faster way to work in Animation Creator: save your characters and styles once and reuse them, see the whole setup on one page, and keep track of everything that is generating across Moduvo.

    • New

      Asset Library

      Save Characters, Styles and Environments, share them with your workspace and add them to any video with Pick from Library.

    • New

      Pro Mode

      Show all wizard steps on a single page for faster setup. Moduvo remembers your choice.

    • New

      Active generations

      A floating indicator shows everything that is generating right now, from videos and images to upscales and voice, with up to 5 jobs running at once.

    • Improved

      Enhance with AI, side by side

      The prompt enhancer adapts to the selected model and shows its suggestion next to your original. Accept it, edit it or keep yours.

  12. Animation Creator

    Seedance 2.0 in 4K, with video and audio references

    The full Seedance 2.0 lineup is now available, including 4K output, and you can guide a clip with reference videos and audio, not only images.

    • New

      Seedance 2.0, Fast and Mini

      Clips from 4 to 15 seconds with optional audio. Seedance 2.0 renders up to 4K, Fast and Mini up to 720p for quicker, cheaper drafts.

    • New

      Video and audio references

      Combine up to 9 reference images with up to 3 videos and 3 audio clips to match motion, style and sound.

  13. Image Generator

    Ideogram 4.0 for images with perfect text

    Posters, banners and social graphics where the headline actually reads right. Ideogram 4.0 is now available for creating and remixing images.

    Good to know

    Ideogram released version 4.0 on June 3, 2026. Typography is its main strength, and it ranks among the best models for design work where legible text matters as much as the image.

    • New

      Ideogram 4.0

      2K images in 15 aspect ratios, from 1:1 to 3:1 and 1:3, with art style presets.

    • New

      Ideogram 4.0 Remix

      Start from one of your images and transform it while keeping its composition.

    • New

      Three speeds

      Turbo for quick drafts, Default for everyday work and Quality for the final version.

  14. Image Generator

    Image Upscaler: up to 6x sharper images

    Turn small or older images into print-ready files. The new Image Upscaler enlarges images up to 6x and adds real detail instead of just blowing up the pixels.

    • New

      2x, 4x or 6x upscaling

      Upload a JPG, PNG or WEBP and enlarge it, with the maximum input size shown for each scale.

    • New

      5 models for different images

      High Fidelity for photos, Low Resolution for tiny sources, CGI for renders and Text Refine for screenshots and packaging.

    • New

      Face enhancement

      Optional face restoration with strength and creativity sliders, plus a history of all your upscales.

  15. Image Generator

    Image-to-image keeps working in the background

    Long image edits no longer depend on your browser tab staying open. Start a generation, move on, and find the result in your history.

    • Improved

      Background processing

      Generations finish on our side even if you close the tab or lose connection. You are only charged if the image is actually delivered.

    • Fixed

      All variants show up

      When you ask for several variants, each one appears as soon as it is ready, with a clear count of how many are still coming.

    • Fixed

      Downloads

      Downloading single images and whole variant groups works reliably again.

  16. Animation Creator

    Aleph 2.0 video editing and Grok Imagine 1.5

    Change an existing video with a single sentence. Aleph 2.0 edits only what you ask for and leaves the rest of the shot untouched, and Grok Imagine 1.5 adds another fast image-to-video option.

    Good to know

    Runway released Aleph 2.0 on May 21, 2026 as its flagship video editing model. It is built for swapping products, backgrounds, lighting or style in real footage. xAI made Grok Imagine Video 1.5 generally available on June 17, 2026; Moduvo offered the preview a few days earlier.

    • New

      Aleph 2.0

      Upload a 2 to 30 second clip and describe what should change: replace a product, restyle the scene, add or remove objects.

    • New

      Reference frames

      Pin up to 5 reference images to specific moments in the video to control exactly how the edit should look.

    • New

      Grok Imagine 1.5 (Preview)

      Bring an image to life in 1 to 15 seconds, in 480p or 720p.

Want the platform for your team?

Tell us who would use it and for what. We'll suggest a setup and a price.

Book a call(opens in a new tab)

Prefer email? vit.bilinec@moduvo.app