Backlog

Next up

In Progress

Done

Feature Requests

Anyone interested in talking about the button actions?I was thinking it might be nice to have more options.Long hold: power on and off(I accidentally turn it off when with double tap sometimes)Single tap: muteDouble tap: process conversationIt shouldn’t be too hard to make each of these options select able to different actions, but I’m no programmer so I can only suggest and hope you geniuses can make it work.Cheers.

Under Consideration

Omi CV1 Device

Acknowledged

Button Actions: custom and ideas

An Anonymous User

Feature Description: Omi will introduce a contextual reminder feature that adjusts reminders and schedules based on real-time external factors such as weather, traffic conditions, and location. This feature will help users make better-informed decisions by dynamically adapting their reminders.Use Cases:<ul><li>A user says, "Omi, wake me up at 7:00 AM for work," and Omi adjusts the wake-up time based on real-time traffic conditions.</li><li>A user says, "Remind me to take an umbrella tomorrow," and Omi activates the reminder only if rain is expected.</li><li>"Omi, remind me to go to the gym," but if it's raining, Omi suggests an indoor workout alternative.</li><li>When the user asks, "What’s the weather like today?", Omi provides suggestions based on planned activities.</li><li>A user requests, "Omi, let me know when I should leave for my meeting," and Omi factors in traffic data to give a precise notification.</li></ul>Benefits:<ul><li>Optimizes user schedules based on contextual data.</li><li>Saves time by preventing unnecessary delays.</li><li>Helps users make better decisions regarding outdoor plans.</li><li>Enhances user experience with personalized and intelligent reminders.</li></ul>Potential Challenges and Risks:<ul><li>Ensuring reliability of traffic and weather data.</li><li>Preventing excessive schedule modifications that may disrupt user routines.</li><li>Performance optimization for real-time data processing.</li></ul>Implementation on Android and iOS:<ul><li>Android can use "Foreground Service" for continuously updating contextual reminders.</li><li>iOS can leverage "Background App Refresh" and "Location Services" for smart notifications.</li><li>Real-time weather and traffic updates can be fetched via API integration.</li><li>Cloud synchronization ensures reminders remain consistent across devices.</li></ul>Proposed Working Mechanism:<ol><li>The user sets a reminder for a specific time or event.</li><li>Omi analyzes weather, traffic, or other contextual factors.</li><li>If necessary, Omi updates the reminder to optimize the user’s plan.</li><li>The user receives a notification about the updated reminder.</li><li>Users can customize sensitivity levels for contextual recommendations.</li></ol>

Smart Contextual Reminders

Ibrahim Albayrak

I want my OMI to recognize personal patterns over a period of time based on stuff I say, experience, etc. And then I want there to be a "Hub" where I can see these patterns and provide feedback (helpful/not helpful).For example, maybe I talk way too much when I have 1:1s with co-workers. I want my Pendant to tell me that. Or maybe my mood tends to be worse when I wake up before 7am. Or maybe I'm a really good storyteller - I want my Pendant to let me know that. Or maybe it can leverage third party data too...things like weather, stock market, key dates, etc.Basically, give me interesting insights/patterns/correlations in my life that I wouldn't think to recognize or connect.

Quiero que mi OMI reconozca patrones personales a lo largo de un periodo de tiempo basándose en lo que digo, mis experiencias, etc. Y luego quiero que haya un «centro» donde pueda ver estos patrones y proporcionar comentarios (útiles/no útiles).Por ejemplo, tal vez hablo demasiado cuando tengo reuniones individuales con mis compañeros de trabajo. Quiero que mi colgante me lo diga. O tal vez mi estado de ánimo tiende a empeorar cuando me levanto antes de las 7 de la mañana. O tal vez soy muy bueno contando historias: quiero que mi colgante me lo diga. O tal vez también pueda aprovechar datos de terceros... cosas como el tiempo, el mercado de valores, fechas clave, etc.Básicamente, quiero que me proporcione información interesante, patrones y correlaciones de mi vida que yo no sería capaz de reconocer o relacionar.

Muéstrame los patrones ocultos en mi vida.

Show me Hidden Patterns in my Life

Petr Korolev

The point of this device is to use your voice for commands and having a personal note taker. I think when you open the chat, it should automatically start using either the phone mic or Omi mic to talk with Omi to have a conversation. Right now it feels rigid with needing to either type or manually activate the mic each time

Low Priority

Ability to speak to chat through Omi

Maverick Gold

Hi Omi team,I used the in-app export to download my data. After exporting, I deleted my data inside the app. I now want to bring that exported data back into Omi, but there’s currently no way to re-import it.Request Add an Import feature that can take an export file created by Omi and restore it into the app.Why it matters<ul><li>Prevents permanent loss after accidental deletion</li><li>Makes export an actual backup workflow</li><li>Helps when switching phones or reinstalling the app</li><li>Builds trust. Users can safely experiment knowing they can restore</li></ul>Suggested UX<ul><li>Settings → Data → Export / Import</li><li>Import flow shows a preview: date range, item counts, size</li><li>Options:<ul><li>Merge with existing data</li><li>Replace existing data</li><li>Import into a new workspace/profile (optional)</li></ul></li></ul>Safety checks<ul><li>Confirm before overwrite</li><li>Warn if file format version differs</li><li>Clear error message if file is corrupted</li><li>Local-only import is fine. Cloud optional</li></ul>If helpful, I can share the export file format details and the exact steps I took.Thanks, Jonah

In Review

High Priority

Android

Import previously exported data back into the app

I would really like the LED to display mute status. Nothing crazy, just a simple colour change between the two states, or the LED switching off when muted (a colour change would be preferable). Muted, connected: yellowUnmuted, connected: blue Muted or unmuted, disconnected: red But that’s just a suggestion.

Mute indicator

dave.io

I’m really excited for my consumer Omi, but I have a couple suggestions for features after looking through the app. Forgive me if these already exist, but I am going off of what I’ve seen.First, the ability to mark a prompt as “repeat daily” would be a huge time saver for people who want daily advice, or check ins, on any given topic or practice. This feature could grant users the ability to gain a different perspective on their daily habits, or it could just make the bot say “Pizza time!” every day at noon.Second, a “Custom Word Dictionary” could enable users to add slang or uncommon business/hobby words to their bot’s vocabulary. This could enable anyone to get tailored advice on conversations filled with uncommon words, or strings of uncommon words.Those are all the suggestions I have for now! Thank you for reading, and please have a good rest of your day/night. If anyone has any questions or corrections, please feel free to share.<ul><li>Draven, USA</li></ul>

Automatic Daily Prompts

fritzkitty

Feature Description: A new feature will be added to Omi that allows users to set reminders for a specific period. If the user does not specify a duration, Omi will default to reminding them for one hour. Additionally, when a reminder is created, Omi will immediately send a notification to confirm its creation.Use Cases:<ul><li>The user says, "Remind me of the meeting tomorrow at 1 PM," and Omi sets the reminder accordingly.</li><li>The user says, "Remind me for 3 hours to take work breaks," and Omi reminds them at regular intervals for 3 hours.</li><li>The user simply says, "Remind me at 3 PM," and Omi assumes a default reminder duration of 1 hour.</li><li>The user says, "Remind me in two days at this time," and Omi schedules the reminder accordingly.</li><li>Upon creating a reminder, Omi immediately sends a notification confirming "Reminder created."</li></ul>Benefits:<ul><li>Helps users keep track of important tasks and events over a specified period.</li><li>Defaults to a 1-hour reminder duration if no specific duration is provided.</li><li>Provides immediate confirmation when a reminder is created.</li><li>Enhances time management in both personal and professional life.</li></ul>Potential Challenges and Risks:<ul><li>NLP algorithms must accurately interpret user input regarding time expressions.</li><li>A strategy for prioritizing overlapping reminders must be developed.</li><li>A confirmation mechanism may be needed if the user provides incomplete or unclear instructions.</li></ul>Implementation on Android and iOS:<ul><li>Android can utilize Foreground Services for continuous reminders.</li><li>iOS can use Local Notifications to repeat reminders at specific intervals.</li><li>Cloud synchronization can be implemented to manage reminders across multiple devices.</li></ul>Proposed Working Mechanism:<ol><li>The user tells Omi, "Remind me tomorrow at 2 PM."</li><li>Omi creates the reminder and immediately sends a "Reminder created" notification.</li><li>At the specified time, Omi notifies the user via voice and text.</li><li>If no duration is specified, Omi continues to remind the user for 1 hour.</li><li>The user can stop the reminder by saying, "Turn off the reminder."</li></ol>

Advanced Reminder: Scheduled & Timed Alerts

<h3>1. Feature Description:</h3>Context Awareness Core is a built-in intelligence module that analyzes the user’s situational data (location, time, environment, hardware status, calendar) and allows Omi to make smarter, context-aware decisions. It provides a background intelligence layer for other systems, enabling them to adapt based on real-world conditions.<div data-type="horizontalRule"><hr></div><h3>2. Use Cases:</h3><ul><li>At home → relaxed mode</li><li>At work → focus mode</li><li>Late night → silent notifications</li><li>On the move → reduce voice prompts</li><li>During meetings → silent assistance and note summarization</li><li>With headphones → enable focus sessions</li><li>Airplane mode → emergency fallback behavior</li></ul><div data-type="horizontalRule"><hr></div><h3>3. Benefits:</h3><ul><li>Smarter timing and interaction</li><li>Adaptive notifications</li><li>Minimizes distractions</li><li>Enhances other systems’ decision-making accuracy</li><li>Makes Omi feel genuinely aware of surroundings</li></ul><div data-type="horizontalRule"><hr></div><h3>5. Android &amp; iOS Feasibility:</h3><ul><li>Android: Flexible access to sensor and environment data</li><li>iOS: Restrictions on background and contextual tracking</li><li>Solution: Combine mobile app (calendar, system) with wearable device (audio, movement, proximity)</li></ul><div data-type="horizontalRule"><hr></div><h3>6. Suggested Working Principle:</h3><ol><li>Data Collection from multiple sources</li><li>Context Profile Generation:Using the above inputs, Omi generates a live context profile that defines the user's current environment and mental mode.For example:<pre><code>Location: Home | Time: 22:30 | Ambient: Quiet | Headphones: No
→ Context: User is in wind-down or relax mode</code></pre>This profile becomes a foundation for adaptive decision-making.</li><li>Adaptive System Tuning for notifications, suggestions, responses</li><li>Periodic Refresh or on-change update every 10–15 mins</li></ol>

Contextual Awareness

Feature Definition: Omi should have the ability to add event-based reminders. Users should be able to give commands like "Omi, notify me when X event happens," and Omi should track the event's date and notify them accordingly.Use Cases:<ul><li>Movie Release Notification: "Omi, notify me when Dune: Part 3 is released."</li><li>Concert or Event Reminder: "Omi, let me know when The Weeknd’s 2025 tour starts."</li><li>Tech Launches: "Omi, notify me when the new iPhone is released."</li></ul>Benefits:<ul><li>Users won’t miss important events.</li><li>Modular design makes it adaptable for various types of events.</li><li>Enhances user experience and usefulness of Omi.</li></ul>Challenges &amp; Risks:<ul><li>Reliable Data Sources: To determine event dates, Omi should fetch data from trusted APIs such as IMDB, Google Events, Ticketmaster, and Apple/Google Calendar.</li><li>Spam Notifications: Users may add too many events, leading to notification clutter. A filtering system is necessary.</li><li>Customization: Users should have the ability to choose what type of events they receive notifications for.</li><li>Date Changes: If an event date changes, Omi should be able to track and update it.</li></ul>Handling Missing Dates:<ul><li>If Omi cannot find an exact event date, it should notify the user: "I couldn't find an exact date for this event. Would you like me to keep looking or suggest alternative dates?"</li><li>Omi will check additional data sources and provide an estimated date.</li><li>Users can manually input a date or ask Omi to check again later.</li></ul>Implementation on Android &amp; iOS:<ul><li>Android: Firebase Cloud Messaging (FCM) and Google Calendar integration.</li><li>iOS: Apple Push Notification Service (APNS) and Siri/Calendar integration.</li></ul>Suggested Workflow:<ol><li>User gives a command: "Omi, notify me when X event happens."</li><li>Omi fetches and stores the event date.</li><li>Reminder notifications can be sent before the event.</li><li>When the event occurs, Omi sends a final notification.</li><li>Users can update or cancel the event reminder if needed.</li><li>If an exact date is not found, Omi provides alternatives.</li></ol>

Event Reminder

It would be nice to be able to link both Google and Apple logins to the same account, so you can log in with both.

Planned

Unified Google & Apple & Microsoft login

Mario Suppa

I think it would be helpful if you just had one main to-do list and they all came in from the start the default color that they are now but you could hold the list item down and select a color and give that color a label like work, school, personal similar to how you tag different speakers in transcripts, that way you can sort your to-do list and then maybe hold the to-do list tab at the top and it would drop down the ones that you created and you can pick which one you want to be at the top. I think it also gives you the ability down the road to let the app learn trends of what items go and what list the same way an untagged speaker when you go to tag it it'll give you a suggestion.

Task List w/ Tags & Filters

MooseSoup

Call Recording

Lyondhür Picciarelli (Týrsdag)

It's great to have when I select a part, it tells me which section of the transcription it comes from. I want to read more about a certain part in the summary, but I need to find it in the transcription having to read through a lot of text.like with perplexity; when I hover over or select the text, it gives me the exact source. I would like something similar in the summary so that when I select a part, it tells me which section of the transcription it comes from.<video data-width="100%" data-align="left" data-featurebase-content-filename="Screencast From 2025-06-02 18-26-18.mp4" data-featurebase-content-key="eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJvYmplY3RLZXkiOiIwMTk3MzI0MC0zYTVlLTdkYTctYjc2Ni1kZGQxMzI3OTE1MjAiLCJ0eXBlIjoidmlkZW8tdXJsIiwicmVzb3VyY2VJZCI6IjY4M2UwNDRhZDAxNDNiMmYyNWVhZmUyMSIsInJlc291cmNlVHlwZSI6InBvc3QiLCJvcmdJZCI6IjY3OWFhNDdlNDVhNmU5Y2VmOTI5YTliOCIsImlhdCI6MTc0ODg5NDc5NH0.l3Mmns8gHRN5AZ-PVG2pQzlOavEPeXCp6uwOggC6OOk" src="https://679aa47e45a6e9cef929a9b8.featurebase-attachments.com/c/videos/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJvYmplY3RLZXkiOiIwMTk3MzI0MC0zYTVlLTdkYTctYjc2Ni1kZGQxMzI3OTE1MjAiLCJleHAiOjE3NzIwNzEyMDAsInR5cGUiOiJ2aWRlby11cmwiLCJvcmdJZCI6IjY3OWFhNDdlNDVhNmU5Y2VmOTI5YTliOCIsInJlc291cmNlSWQiOiI2ODNlMDQ0YWQwMTQzYjJmMjVlYWZlMjEiLCJyZXNvdXJjZVR5cGUiOiJwb3N0In0.eOenZ2IivsTlnICpzz8bKrLXMugdZu_8rJa2GvN1TGI/manifest.mpd"></video>

Completed

Chat With Summary

shailendra thakur

Hopefully the right place for this; just an idea for the Omi App.With summaries being pretty constant, l will have 50+ a day.Could a bookmark feature be added, so we can save important summaries.Alternatively, like within ChatGPT itself now, could a folder system be introduced so we can organise the summaries?

Conversations Folders

the_ai_lecturer

You can link geocoordinates to the timestamps of your transcript. This gave the LLM (Gemini, in my instance) the context that if I'm at home, (or a residence), the chance that I have a panel of people debating the news, or me going to Pandora wouldn't make sense, and it's most likely media. Conversely, it knows if I'm at a theatre and as such, knows the majority of what's going on is most likely a movie. Also if Inoticed me moving large distances it recognized it as travel So it would say "while travelling to x" and contained the transcript, again noting that there's a higher chance of podcasts or music about to be played.If the text sounds more like pe ople talking intensely about a certain topic, and had things like sponsors or an ad break then obviously mark it as such.Worked out really well.

Geolocation

Patrick C.

Speaker diarization currently struggles with accurately identifying speakers, especially in single-speaker conversations where the system should confidently recognize one speaker at all times. Additionally, users should have the ability to tag speakers manually, and those tags should persist across conversations, reducing the need for repeated identification.<h3>Challenges</h3><ul><li>Inconsistent speaker recognition: A user speaking solo should be identified with 100% accuracy.</li><li>Limited diarization accuracy: The current model (Deepgram’s diarization) misidentifies or fails to consistently assign speaker labels.</li><li>Lack of persistent speaker tags: Users must repeatedly tag speakers, as the system does not remember identities across different conversations.</li></ul>

Improved Speaker Diarization & Persistent Speaker Tagging

Salman Mian

From an AI perspective the speech to text quality completely depends on input data quality. There really ought to be an easier way to listen to how 'omi hears you'.

Allow users to hear what Omi hears

kamiirl

Sometimes we may have a summary that spans multiple concepts or conversations, it would be great to be able to find that location in the transcript and split it into two or more transcripts then run the summary again. Or even sometimes maybe just delete something from the transcript, for example on Omi desktop today - I finished a meeting and forgot to stop the transcript and started talking to someone after the meeting in my office, I was planning to share the conversation with my meeting participant but now I have extra information I don’t need in that transcript/summary that is unrelated to the meeting without any way to remove it before sharing. Pairs well with <span data-type="mention" class="dark:bg-border bg-[#E9ECF1] px-1 py-0.5 text-sm shadow-none rounded-md" data-id="6883d4664153bdcf293b7693-post" data-label="Add ability to edit Conversations to enable a better shared experience">@Add ability to edit Conversations to enable a better shared experience to edit some things before sharing and @Option to Manually Combine Summaries because sometimes a longer conversation is split up into multiple conversations and you may want to bring them back together into one summary.

Merge Conversations

Nathan Sudds

Ability to detect the language automatically and transcribe it to the language of your choosing while also adding more languages

Automatic language detection

buls01

Red light is on and maybe I'm doing something wrong but I can't get it to connect. Worked for a few days. Charge each night then it just stopped connecting.

Das rote Licht leuchtet und vielleicht mache ich etwas falsch, aber ich bekomme keine Verbindung zustande. Es hat ein paar Tage lang funktioniert. Ich habe es jede Nacht aufgeladen, dann hat es einfach aufgehört, sich zu verbinden.

Rotes Licht Fehler

Red light Error

Clark Rector

Sometimes, a single conversation results in multiple fragmented memories due to backend latency or system behavior. This can make it difficult to follow the full context of a discussion.A solution would be to allow users to select multiple memories and manually combine them into a single entry. This would merge the transcripts and retrigger memory creation, producing a more cohesive summary.Challenges<ul><li>Multiple memories from a single conversation can clutter the list and cause context fragmentation.</li><li>No way to merge related memories that should belong to the same event.</li><li>Loss of clarity when reviewing past conversations due to unnecessary separation of memory entries.</li></ul><h2></h2><h2>Benefits</h2>✅ Provides better organization by reducing memory fragmentation.✅ Ensures that conversations remain contextually intact.✅ Gives users more control over how memories are structured.✅ Allows for reversibility by keeping links to original memories.Would love feedback on this! Thanks! 🚀

Manual Summary Merging

Since Omi Desktop’s recording function does not work for many Mac OS users, I suggest we add support for the Omi hardware pendant and glasses. This would allow me to use the Omi app on my computer.