# Frigate with MemryX on Home Assistant OS **URL:** \ **Category:** Frigate\ **Tags:** docs\ **Created:** [July 17, 2026, 9:28pm UTC](https://community.memryx.com/t/frigate-with-memryx-on-home-assistant-os/228 "2026-07-17T21:28:59Z")\ **Posts on this page:** 3\ **Page:** 1
**Author:** ![lindsay.morel](https://avatars.discourse-cdn.com/v4/letter/l/59ef9b/32.png) [@lindsay.morel](https://community.memryx.com/u/lindsay.morel)\ **Post date:** [July 17, 2026, 9:28pm UTC](https://community.memryx.com/t/frigate-with-memryx-on-home-assistant-os/228/1 "2026-07-17T21:28:59Z")
# Overview We are pleased to announce that MemryX detector support is now available for Frigate users working with Home Assistant OS (HAOS)! The purpose of this guide is to provide resources for building a custom HAOS image with MemryX support, enabling users to properly configure Frigate with the MemryX detector. Home Assistant OS is a Buildroot-based operating system whose native source code does not include MemryX drivers or runtime. As such, to obtain an image with the MemryX support needed for Frigate, users must build a custom image [from source](https://github.com/home-assistant/operating-system) using the [memx\_haos](https://github.com/memryx/memx_haos) repository as an overlay. # Instructions This guide will provide a high-level overview of the steps needed to build the image and configure Frigate (Full Access) on HAOS. For detailed, step-by-step instructions, please refer to the [`README.md`](https://github.com/memryx/memx_haos/blob/main/README.md) file in the MemryX repository. The procedure has been validated only when followed exactly as documented, so do not skip, reorder, or modify any steps. **Important Note: The [`memx_haos`](https://github.com/memryx/memx_haos/tree/main) repository has been validated only for the Home Assistant OS `generic_x86_64` image and the Frigate Full Access add-on. Other HAOS targets (aarch64, Raspberry Pi, Yellow, etc.) may require additional package changes and have not been tested.** ## Clone the source code repositories. To begin, you will need to clone the [HAOS source code](https://github.com/home-assistant/operating-system) and the [MemryX overlay](https://github.com/memryx/memx_haos) into a common folder. ## Apply the MemryX overlay. Use the commands in the `memx_haos` [`README.md`](https://github.com/memryx/memx_haos/blob/main/README.md) to apply the MemryX overlay to your HAOS source code using the provided bash script. This will apply all MemryX components needed to support Frigate with MemryX on your HAOS image. ## Build the image. Use the commands provided to build the custom HAOS image. You should have at least 40 GB of free disk space for the build to complete successfully. The process will take approximately one hour to complete and will vary based on your system specs. ## Flash the image. The built image can be found in the `output/images` folder with the extension `.img.xz` please flash the image to a boot-able USB or SSD of your choice using either Balena Etcher or `dd`. ## Boot the image. Install the flashed USB/SSD into your preferred x86\_64 system and boot. Once startup completes, use the commands in the guide to verify the presence and functionality of all MemryX components. ## Install Frigate on HAOS from the MemryX fork. Add the [MemryX-configured Frigate HAOS add-on repo](https://github.com/memryx/frigate-hass-addons) URL to HAOS (Settings → Apps → Install App → Repositories → Add). After adding the repo URL, navigate to the app installation page and install “Frigate (Full Access)”. Once installation completes, start the app. ## Configure Frigate as usual. Add and enable your camera(s) first. Then, navigate to the Frigate configuration editor to add the MemryX detector and specify a detection model of your choice. Save the config file and restart the application to begin use of the MemryX detector. # Conclusion If you are interested in using Frigate with the MemryX detector on Home Assistant OS, please refer to the [`memx_haos`](https://github.com/memryx/memx_haos/tree/main) repository to guide you through the process. This post is meant to serve only as a high-level summary of the required actions. As always, the MemryX team is here to support! ---
**Author:** ![Karen\_Nakamura](https://yyz2.discourse-cdn.com/flex008/user_avatar/community.memryx.com/karen_nakamura/32/77_2.png) [@Karen\_Nakamura](https://community.memryx.com/u/Karen_Nakamura)\ **Post date:** [July 21, 2026, 9:07am UTC](https://community.memryx.com/t/frigate-with-memryx-on-home-assistant-os/228/2 "2026-07-21T09:07:59Z")
First, this is awesome. I have my frigate in a TrueNAS VM so it doesn’t apply to me, but I’m still happy to see active work. The latest Memryx SDK and firmware still doesn’t work with the latest frigate, right? Is that going to be fixed soon? ---
**Author:** ![abinila\_siva](https://avatars.discourse-cdn.com/v4/letter/a/c67d28/32.png) [@abinila\_siva](https://community.memryx.com/u/abinila_siva)\ **Post date:** [July 21, 2026, 2:57pm UTC](https://community.memryx.com/t/frigate-with-memryx-on-home-assistant-os/228/3 "2026-07-21T14:57:46Z")
Hi @Karen_Nakamura, thanks for your continued interest in our Frigate integration! At the moment, the latest Frigate release supports MemryX SDK and firmware `v2.1`. We’re planning to update the integration to support our upcoming SDK `v2.3` once it’s available, but we don’t have a timeline for that yet. In SDK `v2.3`, we are adding an MX3 runner for face-recognition models, allowing the MX3 to be used for both object detection and face recognition simultaneously. As part of this work, we will move the integration from `local mode` to `shared mode`. This will: 1. Allow multiple models, such as detection and recognition, to run simultaneously. 2. Remove the requirement to run the Frigate container in `privileged mode`. 3. Loosen the SDK version-matching requirements between the container and the host. We’ll continue working with the community and will post an update here as soon as support for the newer SDK is available. You can also take a look at this post, which provides more details about our upcoming release: > [@Frigate on TrueNAS](https://community.memryx.com/t/frigate-on-truenas/214/2): > > Awesome work, glad to see you got it running on TrueNAS! tada Regarding privileged, we are working on an improvement to our Frigate integration, which among other things will remove the need for privileged. We’re adding an MX3 runner for face recognition models, so it can be used for both the detector and recognition simultaneously. We’ll switch from [‘local mode’ to ‘shared mode’](https://developer.memryx.com/runtime/usage/shared_local.html), which will: Allow multiple models (detection + recognition) to run simultaneously. Remove the need for privil… Thanks again for your patience and support!