Feature Requests

Para este modelo que es muy creativo, 1800 caracteres es limitar la creatividad del modelo, es lo suficientemente inteligente para leer y realizar escenas complejas, mínimo pondría 2200 caracteres. Gracias!Lourdes @iapulse

A song creator that makes custom backing tracks from audio and text prompts. You can choose style, tempo, genre. You can choose a voice to sing it, or just use your own. Add a recording and it makes a track for that song. Add lyrics and it makes a song from it. You can even give an idea and let it do the rest. There’s also a stem splitter and vocal remover. You can even use third-party models like Suno.

I would need a presentation mode in Boards to show my work to my boss without having to show my browser tabs or anything. Simple presentation. Maybe also a Zoom-Tool, to circle certain areas to quickly zoom into specific topics. Like a Mindmap i’d like to have clusters of information, to which i can quickly refer via zoom areas within a presentation mode.

I normally work with multiple ideas within the same board, and one feature I really miss is the ability to group objects such as images, text, lines, and arrows. When working on larger mood boards, I often need to reorganize layouts, compare directions, and move complete sections around.Without grouping, I have to select and move each object individually, which can be time-consuming and I risk accidentally missing elements.A grouping feature, similar to what is available in InDesign and Illustrator, would make it much easier and faster to organize content, keep related elements together, and have a clear overview while working. Warmly, Emmalie Jean

can’t ganerate

Add 'Copy prompt' feature inside Firefly interface integrated within the applications. Copying and regenerating images with different references is a common use case.

Totalmente incorreto. Não há fundo para chroma key. Há uma sobreposição das imagens o que confunde o visual e impede o efeito de ondas. Péssimo trabalho

As we all know, Adobe Firefly now has this "Open in Photoshop Web" feature. But honestly? It just opens the generated JPG/PNG as a single flat layer in Photoshop Web, which is kinda useless. Meanwhile, if you look at how AI is evolving, ChatGPT or Gemini can already export generated images straight to Canva as fully layered, easily editable files (shoutout to Canva’s magic layer feature, it's a lifesaver).So, as the ultimate leader in creative software, Adobe really needs to step up and add a similar feature. We don't want Adobe—the literal king of creative apps—losing its crown and falling behind Canva and its magic layers.I really hope the team considers this feedback. It would help millions of Adobe users out there and, honestly, stop people from canceling their subscriptions and jumping ship to Canva..Here’s the screenshot from Photoshop Web: Canva Magic ​​​​​Layer, it is not perfect but very helpful so we not creating from scratch:  

I am finding the limitation of the characters very difficult to get to the Image I desire. In fact it's Miles away.

En generar voz, quiero reproducir un acento Español (Argentina) pero el habla resulta un español de España, no se corresponde con la realidad del habla de la regiónEs muy importante lograr un habla natural. En mi caso, intento hacer una locución que suene realmente Argentina, pero habla como española pronunciando zetas y con una entonación que no resulta creíble.

The current model of taking credits that a user has paid for and not providing any useful footage is not acceptable commercially.This might have been the case in the early days of Firefly but not now on a fully released service.If there is concern about users downloading videos (from browser tools) without payment have you considered the same model that Music Libraries use - to ‘Watermark ‘ end result before it is paid for then downloaded?.Why not Watermark a video (generated from text or first frame image) in a way that allows generated content result to be verified, but not used as the Watermark would prevent that?.Perhaps the watermark may need to randomly change throughout the video to make this more secure?.Once a user is happy with the result and on payment of your credits,  the final end footage/image may be downloaded correctly.Arguing that server time needs to be paid up front is no longer going to be accepted by users as they will use competitors instead and your user base wi

Add the model seedance 2.5

Adding pages to the boards, similar to how Figma uses pages in their design files, could help section out the creative further.Example use cases:All the content generated for a Round 1 of Creative Development could live in a page. Then Round 2 would have its own page. Batch photography creation could be segmented across pages. Lifestyle with product on one page, lifestyle without product on another, product imagery on another, etc.Essentially, giving the board even more longevity and collaboration instead of only relying on the sections tool or manually organizing.

Would it be possible to remove, or significantly increase, the token limit for prompts in Adobe Firefly?The current limitation severely restricts the usability and creative potential of the tool. Specifically, when working with Seedance, this limit becomes a major bottleneck. Seedance can easily process very long and detailed prompts without any issues, yet the Firefly platform automatically blocks them once they reach an arbitrary length limit.Having this restriction on the platform side when the underlying workflow can handle it perfectly fine is counterproductive and limits advanced use cases. Removing this cap would vastly improve the user experience and allow for much greater creative control.Thank you for considering this request.

eh comprado mi suscripción de firefly y no me deja subir mis imágenes para generar videos con mis propias imágenes de referencia, no carga ninguna imagen

Why can’t I write long prompts? Why is there a limit on prompt length? I have a paid subscription, but the number of words is limited in Adobe Firefly. Why is that? I’m even thinking about switching to another AI just because of this. What’s the problem? How can I remove this limitation? I need to write a more detailed prompt, but I can’t because of the restrictions!

Runway Gen 4.5 should have 9:16 not just 16:9. Please work to make that possible, I have to get a runway account to generate 9:16.

1

It would be incredibly helpful if Adobe Firefly allowed users to select very specific parts of an image with more precision and change or remove them using simple text prompts. Improving the brush selection or adding automatic object detection would make background removal and object replacement much faster and more accurate for digital marketing content."

Da siempre error la generacion de video si no llego a tener la prueba gratis pago para no poder generar videos alguien da una solucion?

Each time I go to Generate Video, usually via Lightroom Classic’s Generate Using Firefly menu commands, the Resolution is 720, Audio is disabled, the Seed is blank.And I then change them to 1080, enabled Audio, often type in the same Seed that I used last time (because I’m doing a few similar pictures).Sometimes I forget, and burn a few credits, other times I have to remember to waste a few moments changing the settings.Why can’t these settings be sticky?

no puedo animar imagenes en kling 3.0 en mi prueba gratuita  

Hello Adobe team. I think there is an issue with Firefly video generation using Seedance 2.0 / Seedance 2.0 Fast with image references.Text-to-video works normally, but any image-to-video generation fails. I tested multiple reference images, including an original character, a screenshot version, and a normal photo of a cat. All fail with the same error.Error message:“Gemini classification API returned HTTP 404. Model gemini-3.1-flash-lite-preview was not found or the project does not have access to it.”This happens on both the iPhone Firefly app and the desktop browser version. 

Firefly is great at expanding images, but if you start with a horizon that is slightly crooked, the results are often not good. Asking the AI to straighten the horizon often leaves an offset. It would be great to have an option to straighten the image manually

ele trava sozinho