Udio Live Q&A Recap: Mobile Apps, Song Editing, 2.0

Official Udio Live Q&A Summary

Session Overview and Atmosphere
Mobile App Development
Model and Feature Development
Current Feature Discussions
Community and Social Features
Miscellaneous and Fun Moments
Summary of just the team commentary
Summary of just the side chat
Udio Development Status
Future Vision

---

Session Overview and Atmosphere

The Q&A session took place on Discord with Udio team members Adam, Johannes, Eshan, and Ian participating while dozens of community members actively commented in the chat.

The session began with casual conversation among users while waiting for the official start, with many joking about the background "elevator music" playing during the wait.

Throughout the session, the chat remained lively with users reacting in real-time to the team's responses, often with emojis, jokes, and follow-up questions.

Mobile App Development

iOS and Android Apps:
Adam confirmed they're developing an iOS app first, followed by Android.
When asked about timing, Adam managed expectations by explaining they're still finishing internal development before moving to testing.
Several users in chat expressed excitement, with @skich mentioning they had already DMed Johannes about iOS testing.
Feature Parity:
Adam noted the mobile apps won't have feature parity with desktop at launch.
Team is focusing on making the discovery experience more fluid on mobile.
Many chat participants expressed they primarily use Udio on mobile already.

Model and Feature Development

Song Editing Improvements:
The team confirmed they're actively working on allowing users to cut sections from the middle of songs.
In chat, @ixus_82752 commented: "Yeah a lot of times I'd love to erase a literal 4 measure section of song to condense and that is currently an impossibility."
Users expressed excitement about interior cropping and the potential to insert sections in the middle.
Voice and Melody Consistency:
Multiple users repeatedly requested the ability to save and reuse voices across songs.
@filmmakerzach commented: "I want consistent voices so bad."
Eshan acknowledged this challenge, explaining they're working on isolating different dimensions (vocals, style, instruments) for consistency.
Model Updates ("2.0"):
When asked about a "2.0" model, Eshan couldn't confirm specifics but mentioned they're working on speed, controllability, and flexibility.
The chat erupted with excitement at the mention of 2.0, with users playfully asking "when 2.0" and joking about version numbers.
Johannes later quipped, "when we release 2.0, people will still say they like 1.0 more," which received laughing responses from chat participants.
Melody to Full Arrangement:
The team acknowledged high demand for a feature allowing users to upload simple melodies and transform them into full arrangements.
Team expressed excitement but maintained they have high standards before releasing.

Current Feature Discussions

In-painting vs. Edit:
Several users debated the merits of the in-painting feature versus the newer edit feature.
@riptanion commented: "Replace is great for redoing instrumental parts, but if you're fixing vocals, inpaint is still king."
@reaperscreepers added: "+1 for inpainting because you can edit multiple sections at once."
Trash Bin Feature:
Johannes mentioned he built the trash bin feature, which received enthusiastic appreciation in the chat.
@kitzunekaze commented: "Trash bin has saved a couple of my songs. Thank you Johannes!"
@allinfavorai joked: "Moment of silence for all the pre-trash bin accidental deletes."
Rendering Options:
Discussion of different rendering modes (turbo/normal/eco) similar to other AI platforms.
Team is exploring preview options where users could generate multiple shorter samples.

Song Discovery and Promotion:
Ian mentioned they're looking to improve the home screen for better discovery experience.
Eshan acknowledged they haven't prioritized recommendation algorithms yet but plan to develop personalized recommendations.
In chat, @skich mentioned having "converted probably close to 50 people" to Udio and wished for affiliate links.
User Profiles:
@robertejames requested: "Please add the ability to pin your own fave 3 songs on your profile."
Johannes confirmed this was "on the list."
Chat participants joked about creating a MySpace-like experience, with @johannes_udio joining in: "we should remake myspace/tumblr personalization with a song that plays when you go to the profile."
Staff Picks and Community Content:
Johannes mentioned they still add user-submitted content to spotlight features.
Chat messages revealed that some users' staff picks had been accidentally deleted before the trash bin feature existed, with @telerune and @skich sharing an anecdote about a song called "Vanish" that ironically vanished.

Miscellaneous and Fun Moments

Adam's Voice: Several users complimented Adam's speaking voice, suggesting he should host a podcast, with @jewishguy613 commenting: "Adam's voice is peak podcast, I could listen all day."
Community Size: Adam mentioned a joke statistic of a user having "42,000 songs" which was later revealed to be closer to the actual number of songs on the platform.
Captcha Confusion: One user (@cybergetixsx) complained about "pop-up ads," which other users and Johannes clarified were actually just captchas.
Credits and Sharing: Several users suggested implementing gift credits systems to share with friends.
Closing Appreciation: The session ended with numerous thank you messages from community members expressing appreciation for the team's work and the Q&A format.

Summary of just the team commentary

Feature Development and Roadmap

Song Editing Features: The team is actively working on allowing users to cut pieces from the middle of songs, though currently users need to use external DAWs to accomplish this.
Mobile Apps:
iOS app is being developed first, followed by Android
Apps are in internal development, will go to testing, then release
They won't have full feature parity with desktop at launch
Mobile apps will focus on improving discovery and consumption experiences
Melody to Full Arrangement:
The team acknowledges high demand (150 votes) for a feature allowing users to upload simple melodies/humming and transform them into full arrangements
Similar to but distinct from the current "remix" feature
Team is excited about this but maintains high standards before releasing
Model Updates:
Team is working on improving speed, controllability, and flexibility
Can't confirm specifics about a "2.0" model or timelines
They explained that uncertainty about release dates stems from doing revolutionary work with unpredictable outcomes

User Experience Improvements

In-painting Access: Working on making the in-painting feature more accessible for users with large libraries
UI and Discovery:
Planning improvements to the home screen and discovery to help creators get their songs noticed
Working on a personalized algorithm for song recommendations
Following section on the website may be redone
Rendering Options:
Considering different rendering modes similar to other AI platforms

(turbo/normal/eco)

Exploring preview options where users could generate multiple shorter samples before committing to a full generation
Model Transparency:
Considering ways to expose the model's "reasoning" to help users learn how to craft better prompts
Challenge is determining which aspects of generation are reproducible vs. random

Technical Insights

Bug Fixes: The team held a "bug bash" to improve desktop and mobile usability
Original Model: Confirmed the v1.0 model hasn't changed in months, despite some user perceptions
Consistency Features:
Working on ways to maintain consistent vocals, instruments, and musical sections across generations
Challenge is isolating different dimensions like vocals from style from instruments
Trash Bin Feature: Highlighted as a particularly impactful feature that saved users' work

Business and Community

Funding: Not accepting individual investments at this time
Hiring:
Currently focused on engineering positions
Brought on two new engineers recently
Recently brought on a technical writer
Encourages fans to watch udio.com/careers for updates

Future Vision

Team members described what they want users to say about upcoming features:
Ian: "This is something new" - referencing a visual way to interact with song creation
Eshan: Wants users' friends to say "I need to save this to my library and stream it"
Ian mentioned exciting developments coming in the next two months
A suggestion was made for a community challenge where users would create songs based on topics from friends and record their reactions

Summary of just the side chat

General Discussion and User Experience

The session started with casual conversation among users greeting each other while waiting for the Udio team members to join.
Users discussed the background music playing during the Discord session, with some joking it might have been made in Udio.
Adam, Johannes, Eshan, and Ian from the Udio team participated in the Q&A.
Users expressed appreciation for the trash bin feature, which allows recovery of accidentally deleted songs.
There was discussion about user interface improvements, with some users mentioning they only recently discovered features like selecting timeframes in their library.

Feature Requests and Discussions

Voice Consistency: Multiple users requested the ability to save and reuse specific AI voices across different songs.
Workflow Improvements:
Users requested better control for song generation, including the ability to insert sections in the middle of songs.
Interior cropping was mentioned as a desired feature to edit specific parts of songs.
Users want to name songs before generating them rather than after.
Several users expressed interest in longer song generation (3+ minutes).
Editing Tools:
Discussion about "inpainting" vs. the newer "edit" feature, with many users preferring inpainting for editing vocals.
Users requested better tools for mixing, cutting, and pasting sections of tracks.
Publishing and Sharing:
Request for distribution services similar to DistroKid.
Users wanted the ability to pin favorite songs on their profiles.
Discussion about improving song discoverability and recommendations.
Technical Requests:
NSFW toggle to control language in lyrics.
Ability to download lyrics as .srt or .txt files.
Transcription of uploaded audio.
Better feedback on moderation errors when generations fail.

Udio Development Status

The team mentioned they track changes at feedback.udio.com/changelog.
Johannes confirmed they're working on allowing users to specify how many songs to generate at once (currently fixed at two).
Staff picks are still being added from user submissions, with many submissions received recently.
There was teasing and joking about a potential "2.0" version, but no specific details or timeline was confirmed.
The team acknowledged moderation issues, noting that even they sometimes only get "good/bad" feedback from the moderation system.
Mobile app development is underway, with iOS testing mentioned.

Community

Users discussed creating communities around Udio, with some mentioning they've introduced dozens of new users to the platform.
There was discussion about "SUCC" which appears to be a user-created community or group.
One user mentioned Udio helped them return to making music after a disability prevented them from playing jazz.
A user mentioned creating virtual bands using Udio-generated music.
The Chinese market was mentioned, with a request to add WeChat and Alipay payment methods.

Miscellaneous

Users joked about Adam's voice being suitable for podcasts or ASMR.
Some users mentioned seeing captchas (which one user confused with popup ads).
The session ended with the Udio team thanking everyone and mentioning they may do more collaborative sessions in the future.

Udio Development Status, Community, and Future Outlook Summary

The session combined live team commentary with side-chat interactions, highlighting ongoing work on:

Song editing capabilities including middle-section cutting and interior cropping.
Voice consistency across songs via dimensional isolation and improved control.
Speed, controllability, and flexibility in a potential 2.0 model, with timelines understood as uncertain due to transformative work.
Mobile app development (iOS first, then Android) with a focus on discovery and consumption experiences and with no full parity at launch.
A strong emphasis on community features, discovery improvements, staff picks curation, and user-submitted content.
Rendering modes and preview options to allow multiple shorter samples before full generation.
Model transparency and learning prompts for users to craft better prompts.

This document preserves the flow of the live Q&A, including direct quotes from attendees and team members. It is designed to be a faithful, detailed record of what was discussed, debated, and planned, with sections organized to mirror the original topics. Readers can reference specific sections to understand timelines, feature experiments, and community requests as they stood during the February 25, 2025 session.