Feature Requests

.

Die bereitgestellte Beschreibung ist selbsterklärend.

I was thinking it would be useful to be able to convert the AI jpg into a layered psd file to further edit and alter AI images. This would be of value to also take these files and animate them in AE. The AI generated images are a great start but very limited with a flattened image.

Would be great if i could directly access my images from AEM and link them into my concept board in firefly using Asset link or desktop app. Then when i need to add the images from my firefly concept board to an existing project, i want to be able to select in image in my firefly board and send it to a Indesign or illustrator fiile that already exists in AEM. 

Support for adding Tiff images to Firefly boards. This is necessary for my workflow and is something that was available in XD where my team was previously creating concept boards. Without this I would be forced to create a second set of every image just to be added to the concept board. 

When working to make a short film, the length and quality of the video segments is key to making the overall story seem seamless.  By adding Seedance 2.0 and 2.5 as a video model, it would allow users to create more dialog per video segment to allow for deeper and more cohesive video production without having to splice videos together for a character to complete their sentence of dialog or action sequence.

Hello, I'm new here and for fun I'm converting images and images to videos. I already have some experience from various free AI generation platforms. I understand that AI is still developing and that it sometimes doesn't understand my words completely. Therefore, I would suggest that there be an option to generate a rough preview before the main generation. Because even if I run the generation with the lowest computational power requirements and the result is acceptable, the next video will be generated again according to new algorithms and the result may differ. So either during generation, a number should be generated that will allow the video to generate consistently according to the same sequence. Or it should generate a preview of 1 frame per second.I understand that AI likes to make things up, but if a character is at the same distance in the original, it sticks to the concept, but as it gets closer, it invents interesting, even creepy things and creates characters that don't mat

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

Please bring back the button that was previously located in the bottom-left corner of the Firefly interface, which allowed users to open and close the panel displaying their generated images. This button was very useful for quickly accessing or hiding the generated images panel. If bringing the button back is not possible, please consider keeping the generated images panel permanently visible instead.

some other ai provide there is button in ai assistance to copying all text, add that there it’s more useful. 

Character limits for video generation aren't displayed during the creation step. For instance, while Kling Turbo accepts 500 characters, others accept 1,000. Since we aren't aware of this—and it isn't stated anywhere—we end up producing useless videos. The system should highlight the text in red when the character limit is exceeded, and we should be able to see exactly how many characters we are allowed to use. It’s a very simple improvement, yet somehow it seems to have been overlooked.

The textbox to write and adjust prompts  is way too small for Firefly web. It only shows two lines of text. When doing video prompts it's almost unremarkable.especially when you're using elements. When you want to add an element it always puts the element at the end of the entire prompt and then I have to select and drag it to the right position in the prompt and then finding the right position is really a hassle because you have to scroll through an entire text prompt but you can only see two lines it makes the whole process frustrating and time consuming.I would like to suggest the ability to make the text box bigger like other Adobe apps like for example a slider ready we can make the text box bigger and the results were the videos and generations appear smaller. Right now the text box in the entire interface is less than 15% of the entire surface of the user interface. Which is kind of weird because in the text box there is where all the work really happens and your workspace it's

it should be easy to select and delete

It would be great to have an option to pin the prompt box/dock or disable auto-hide in Firefly.The auto-hiding behavior is frustrating, and dividing image and video options across separate buttons feels counterintuitive and creates unnecessary steps. Combining these functions would be much more intuitive. See the concept below.  

facilitare l'acquisto di nuovi crediti aggiuntivi perché non si trova sul sito dove sia possibile farlo. L'ideale sarebbe poter acquistare pacchetti di crediti una tantum e non in abbonamentonon si capisce dove è possibile acquistare nuovi crediti: sul sito non c’è nulla che lo indichi

The ability similar to that of Apple’s “FreeForm” app, allowing for drawing of annotations on, around, and under placed elements and images. Via mouse or drawing tablet. For mobile versions, styluses or apple pencils

Add a voice cloning feature to the Text to Speech tool that allows users to upload a voice sample and generate speech using that same voice. This would make it much easier to create consistent voiceovers while preserving the original voice's tone, pronunciation, and speaking style. The feature could include consent and verification options to prevent misuse.

Open a footage or multiple footages from Frame.io In Firelfy create a sequence or rough cut Export in XML to open in Adobe Premiere or Resolve on desktop.

normalment je suis sencer pouvoir generer a nouveau du contenue

It would be helpful to use a user uploaded image or video to inspire generated content, or to be able to add backgrounds and objects to the uploaded video like you can in photoshop with a still image

Buenas, Jugando con las opciones de la creación de imágenes de texto a imágen, me ha surgido la idea de tener más opciones respecto a las imágenes de referencia. Cuando subes una imágen, al generar se toma como base únicamente el estilo en el que está la imágen, que está genial, pero no se tiene en cuenta la forma. Podría ser útil que la gente pueda subir bocetos de lo que tiene en mente y a partir de éste y con el texto luego generar la imagén. Como ejemplo, un boceto simple de una mesa redonda que he dibujado y el resultado al generar resulta en una mesa cuadrada en estilo de boceto. Al no especificar la forma de la mesa en el texto, ignora la forma del dibujo o no la interpreta bien, aunque sí mantiene ciertos elementos. Texto usado: mesa de madera antigua deteriorada por el tiempo, en un jardín abandonadoEstilos aplicados primer resultado: Ilustración digitalEstilos aplicados segundo resultado: Foto, Art Decó Un saludo.

Add Ukrainian language interface please.Add Ukrainian language interface in Adobe Reader mobile (android/IOS)

Adobe already ships an MCP connector that lets Claude reach Creative Cloud — asset storage, Lightroom-style adjustments, background removal, vectorise, InDesign merges, video render. It's genuinely good. But the one thing it can't do is the thing Firefly is actually for: generate an image or a video from a prompt.That means my workflow is split down the middle. I plan a sequence with Claude — shot list, prompt text, aspect ratios, continuity notes — and then I have to hand-carry every prompt into the Firefly web UI one at a time, download the outputs, re-upload them, and manually keep track of which frame is which. For a 7-episode vertical micro-serial, that's hundreds of context switches for work the agent could already orchestrate end to end.Higgsfield shipped an MCP server this spring that does exactly this, and it's now the default entry point for a lot of us: connect once, describe the shot, the agent picks the model, fires the generation, polls, and returns the clip in chat. It w

A place where you can file errors and complaints that gives feedback to the developers. Wich would make the report button usefull.

I really don't understand the point of this utterly stupid token limit for prompts with GPT Image 2.5, Seedance or other models, which only serves to make the experience problematic and frustrating for every user trying to work on your platform.This issue has already been widely reported by countless users regarding Seedance, but now it is also happening with GPT Image 2.5.  The previous version was already affected by this restriction, but it could be bypassed by using the edit section instead of generate. With the new model, however, regardless of whether you choose generate or edit, a 1024-token limit is always enforced. This is ridiculously low and makes it difficult to create complex and detailed prompts, as you are forced into a terribly short length.  This makes no sense at all, considering that according to OpenAI's official documentation, these models have no token limit. So I am asking you once and for all: please remove this token limit from Seedance, from GPT Image 2.5, and