Feature Requests

The Enterprise Bottleneck: As enterprise teams scale their use of Descript’s Generative AI, Avatars, and AI Speech features, the final barrier to deployment is the "Uncanny Valley"—micro-jitters, phonetic lip-sync mismatches, and visual anomalies. Algorithms alone cannot clear the final QA hurdle required for strict enterprise brand and compliance standards. The Feature Request: Please build a native integration pathway or secure API webhook designed to route specific AI-generated clips to an external, secure HITL (Human-in-the-Loop) QA layer for anomaly detection before they are approved for the final timeline. The Use Case / Architecture: This integration would allow enterprise users to connect Descript directly to specialized "Human Infrastructure as a Service" (HIaaS) providers, such as Taskmaster Virtual Ally. Taskmaster utilizes a proprietary "2-2-1 Pod" architecture to validate lip-sync, cultural nuance, and visual integrity with strict Zero-Data-Retention protocols. Business Impact for Descript: By opening an integration pathway for specialized HIaaS vendors, Descript becomes instantly more viable for highly regulated enterprise clients. For context, Taskmaster's integration architecture already holds live Production API approval in strict healthcare environments (ModMed) and is fully vetted on the AWS Marketplace. Building a bridge to allow this level of secure human validation natively within Descript would massively de-risk enterprise adoption of your generative tools.

I was trying to onboard and test several video editors for many projects I have in my pipeline. I have learned that the only way to do so is pay for an annual seat for each person. Ok, I understand you are stuck in the mid 2000's with your per seat pricing, but I have also learned that as I'm planning to test them, and many of them would edit 1 video and be done as they won't make the cut or will decide it's not for them, I have ZERO option of adding a different user on the annual seat subscription I just paid for. Is this true? You don't have the capability of reassigning a seat to a new user and are just taking money for an annual subscription which is now just forfeited if an employee leaves? I hope this is not true. I know I had a lot of issues getting the billing set up correctly when I signed up, but these are simple parts of running a business. You guys can handle all this technology to build the functionality, but I am currently really regretting I got involved with Descript as this is literally a train crash for my business. I will now have to move everything to another platform as I'm not throwing money in the garbage can and I'm not going to wait around for you guys to figure it out. I will be staffing up to 10+ editors, but I'm definitely going to have to focus on a more business friendly platform. ****Your chat bot assured me this is the only option so if not, I apologize for the confusion.

The option to edit a word gap has disappeared when I right-click on a gap between words. I'm finding this very annoying and frustrating. I feel like you're forcing me to use the AI Underlord when I don't want to. I have read the instructions and used Descript for a very long time, and it's still not working. I do not want to pay for AI credits to do something I should be able to do manually.

I wish there was a feature that, when I add a b-roll or stock media and click on it while editing, it could display the duration of that stock footage. If I click on the stock media, it should show the duration somewhere, as I want to keep it to 2 or 3 seconds. Currently I have to manually zoom in and count from 21 to 24, then 3 seconds, instead of simply displaying the duration somewhere, either on the hover.

A button that put the selected generated image into a new prompt (like the "Turn into video" button).

Being able to select an actual camera as the 'cutaway' would be super helpful in projects that are a typical of an in-person 3 camera podcast, where the 3rd camera is a wide 2-shot. This could be added to the Camera setup settings where you select the Cutaway camera to be either "split screen" or a track like the other camera option selector. Thanks!

Steps to reproduce: Select a script track video clip. In Properties panel > Animations, add Zoom and pan. Click the keyframes icon (two overlapping diamonds) to enter keyframe edit mode. Select the end keyframe. Timeline shows the blue line with both diamonds correctly, confirming the animation/keyframe is selected. The clip in the Scene Editor shows a purple dashed selection border with resize/move arrows visible. Problem: Dragging the video does nothing. No move, resize, or pan occurs, despite the handles being visibly present and the keyframe being correctly selected. Already ruled out: Layer is not locked (checked Properties > Layers, no lock icon present). Tried both start and end keyframes. Confirmed animation and keyframe selection state look correct in the UI.

I'd love to be able to do interviews with people who don't speak the same languange as me via live translation during recording.

I recently shared the incredibly inconvenient workflow we have to use when we share a project with a VA that does not have a paid account, and why we are not buying a seat for them (let me know if you want the deets, or ask the product manager who recently sent a survey). When sharing a project, why not allow (as an option) the editor we shared with to use AI features, using our AI credits?