Software downloads

This page provides quick access to various Coral software. However, if you're just getting started, you should follow the appropriate setup guide for your device.

Jump to a section:

Note: The Edge TPU Compiler is available only for Linux, in a Debian package.

Mendel Linux link

To flash your board with Mendel Linux, be sure you download the appropriate image from the following tables, for either the Dev Board or Dev Board Mini.

For information about Mendel changes, see the Mendel release notes.

Dev Board link

The following files include everything you need to flash the Dev Board ("Enterprise") with Mendel Linux.

Starting with the Eagle release, there is also a "flashcard" image you can flash to a microSD card, which then automatically flashes Mendel onto the board. For instructions using the flashcard image, read Get started with the Dev Board.

For instructions using the other images, see how to flash the Dev Board.

System requirements:

Release Package Size SHA-256 checksum
5.3 Eagle (Nov 2021) 586 MB enterprise-eagle-20211117215217.sha256
a74f29129877ed196f1f81346e7d85b966e51e7954f244e240bd244832c81c9a 566 MB enterprise-eagle-flashcard-20211117215217.sha256
5.2 Eagle (June 2021) 570 MB enterprise-eagle-20210518200035.sha256
e1e677f52f9d72e6a5bb4e8bed4fbd4683b24cfb052f3e78a63484137696bf9b 565 MB enterprise-eagle-flashcard-20210518200035.sha256
5.0 Eagle (Jul 2020) 547 MB enterprise-eagle-20200724205123.sha256
7ef042fe6b4c5676ab2855146b489870bb034dc4d162537dadaf6dee7ccc7b79 572 MB b25f8774a9dccf09ea9ca7c05371bb7e8b0ec14d3d0692d54dd0ca3de6b3f5a2
4.0 Day (Nov 2019) 541 MB a21205914e46a75ca5bdc9733260a8b6dc4c001513556fd896089bdcd82b596b
3.0 Chef (Apr 2019) 454 MB df733fd22de156f324a1b5fc3d03251a732ddbbf0e34bd54286d42a2e79246e5
2.0 Beaker (Mar 2019) 1.3 GB 85a1db9a6d251a38f34fabf808b4ad3c35d7ab413318bee1d70de48bde776486

Dev Board Mini link

The following files include everything you need to flash the Dev Board Mini ("Excelsior") with Mendel Linux. It also includes the script, which helps put the board into fastboot mode if it fails to boot.

System requirements:

For instructions, see how to flash the Dev Board Mini.

Release Package Size SHA-256 checksum
5.0 Eagle (Dec 2020) 612 MB excelsior-eagle-20201210233645.sha256
5.0 Eagle (factory) 818 MB

MediaTek license agreement

Mendel development tool (MDT) link

Mendel Development Tool (mdt) is a Python-based command-line tool that lets you communicate with a device running Mendel Linux (used on the Coral Dev Board and Dev Board Mini).

You can install it with pip as follows:

python3 -m pip install --user mendel-development-tool

For details, see the MDT documentation.

Debian packages link

If you're using a Debian-based Linux system (including Ubuntu, Raspberry Pi OS, and Mendel), you should install all software libraries with the following Debian packages, instead of using the other pip or ZIP packages.

If you're on a Debian system other than a Coral board with Mendel, first add our package repos:

# None of this is needed on Coral boards
# This repo is needed for almost all packages below
echo "deb coral-edgetpu-stable main" | sudo tee /etc/apt/sources.list.d/coral-edgetpu.list
# This repo is needed for only python3-coral-cloudiot and python3-coral-enviro
echo "deb coral-cloud-stable main" | sudo tee /etc/apt/sources.list.d/coral-cloud.list

curl | sudo apt-key add -

sudo apt-get update

Then you can install packages from the below table like this:

sudo apt-get install <package_name>
Note: When you want to update any one of these packages, use apt-get upgrade to ensure you get all the packages that are compatible with each other.
Package nameDescriptionSupported systems1
edgetpu-compiler The Edge TPU Compiler.

Note: Not available on Coral boards.

  • Debian 6+
  • x86-64
libedgetpu1-max The Edge TPU runtime. (Pre-installed on the Dev Board.) The separate versions affect only USB-based devices to determine the operating frequency: libedgetpu1-max runs at the maximum frequency (500 MHz) and libedgetpu1-std runs at the reduced frequency (250 MHz).

For PCIe-based devices, both packages behave the same because the PCIe driver performs dynamic frequency scaling to adjust the operating frequency based on programmable temperature thresholds.

  • Debian 10+, Ubuntu 16.04+, or
    Raspberry Pi OS Buster and newer
  • x86-64 or Armv8 (64-bit)
libedgetpu-dev The edgetpu_c.h and edgetpu.h header files for running inference with C++.
gasket-dkms Coral driver for PCIe-based Edge TPU devices, such as the M.2 and Mini PCIe Accelerator.
  • Debian 10+ or Ubuntu 16.04+
  • x86-64 or Armv8 (64-bit)
python3-pycoral The PyCoral API.
Includes a dependency on python3-tflite-runtime and libedgetpu1-std (or libedgetpu1-max).
For code examples, see the GitHub repo.
  • Debian 10+, Ubuntu 16.04+, or
    Raspberry Pi OS Buster and newer
  • x86-64, Armv7 (32-bit), or Armv8 (64-bit)
  • Python 3.6, 3.7, 3.8, or 3.9
python3-tflite-runtime The TensorFlow Lite Python runtime API.

python3-edgetpu The Edge TPU Python API.
Deprecated. Instead use python3-pycoral.
This depends on a legacy version of libedgetpu1-std that's not compatible with newer libraries.
  • Debian 10+, Ubuntu 16.04+, or
    Raspberry Pi OS Buster and newer
  • x86-64, Armv7 (32-bit), or Armv8 (64-bit)
  • Python 3.6, 3.7, or 3.8
edgetpu-examples Code examples for the Edge TPU Python API. Saved in /usr/share/edgetpu/examples/.
Deprecated. Instead use python3-pycoral.
This depends on a legacy version of libedgetpu1-std that's not compatible with newer libraries.
python3-coral-cloudiot The Coral Cloud IoT API.

  • Mendel OS or Raspberry Pi OS
  • Python 3.0+
python3-coral-enviro The Coral Environmental Board API.

Includes python3-coral-cloudiot.

1 The "supported systems" are system configurations we currently test, but the packages might work on other similar systems. You can also compile the Edge TPU runtime and Python library for your specific platform using our source code.

Edge TPU runtime link

The following ZIPs include the Edge TPU runtime for macOS and Windows, plus the USB and PCIe drivers required on Windows.

System requirements:

  • Windows 10, or macOS 10.15 (Catalina) or 11 (Big Sur)
  • x86-64 system architecture

If you're using Debian Linux (including Mendel), you should instead install the Edge TPU runtime from our Debian packages.

Note: We periodically update the Edge TPU runtime with small changes such as to improve support for different host platforms, but the underlying runtime behavior remains the same. That's why the following table shows multiple ZIP packages with the same runtime version and different dates in the filename.

You should always use the latest runtime, and be sure your models are compiled with the corresponding version of the Edge TPU Compiler.

Version Package Size SHA-256 checksum
14.3 MB 7f17d85e6f21ec861af3f9168043a7ec351b80e08c9f168013ab68e99410aee2
17.1 MB b23b2c5a227d7f0e65dcc91585028d27c12e764f8ce4c4db3f114be4a49af3ae
17.1 MB 5a490f1cb26a40c6582019a734836239a96f0b06813df98166e0aff219073ac7
17.1 MB 65c6e876030a89a25c011cda7afe5c2820d40ec6f4dee8c86d70c1ae0b8d74bb
16.8 MB f5e1dfd26c37641a4c8eca61f9236e31c355302e5a75c81c690626e777fff67a
16.6 MB 8742039d19715c3274152fe144f1be1f0b3f1dda2e5d605ce90cdac47fc95b28
16.6 MB f98857a43f718d129dd6c1565358a71c8e8008d7947332b488c160c756d5d4ad
12 edgetpu_runtime_20190920.tar.gz
2.8 MB 111c1cef3d8079f5ee9ac8c3cd3ec0822f4a784bc8c8aa4a4affe67d3a6b72a4

Also see the Edge TPU runtime release notes.

PyCoral API link

The following Python wheels are for the PyCoral API.

If you're using Debian Linux (including Mendel and Raspberry Pi OS), you should install PyCoral from our Debian package:

sudo apt-get install python3-pycoral

On other systems, you can install PyCoral with pip:

python3 -m pip install --extra-index-url pycoral~=2.0

Also be sure you have the latest version of the Edge TPU runtime.

Note: If you download a wheel from the following table to install manually with pip install, you must also install the corresponding the tflite_runtime wheel.

macOS 10.15 3.6
macOS 11 3.6
Windows 10 3.6
Linux x86 (64-bit) 3.6
Linux Armv8 (64-bit) 3.6
Linux Armv7 (32-bit) 3.6

Pre-compiled models link

For demonstration and experimentation purposes, you can download and run one of several pre-compiled models that are compatible with the Edge TPU.

Legacy packages link

The following packages are here for archival purposes only.

Release Package Size SHA-256 checksum
2.12.1 (Sep 2019) edgetpu_runtime_20190920.tar.gz
2.8 MB 111c1cef3d8079f5ee9ac8c3cd3ec0822f4a784bc8c8aa4a4affe67d3a6b72a4
8.6 MB 7f5d9abfe429f1ae063259198d665690e655b1459eff3c3bfaf19a8fe62d12ae
2.11.1 (Jul 2019) edgetpu_api_2.11.1.tar.gz
5.1 MB 6067fcc921423c3dc23e1bc12186e42ff954eada50b44826b2167700b50dc69e
1.9.2 (Apr 2019) edgetpu_api_1.9.2.tar.gz
8.3 MB 82324ac028fda707926b756cc97430c8d7f5ffa39940ba74d373de7d9dc90c86
Beta (Mar 2019) edgetpu_api.tar.gz
31.8 MB 52e29f89481e935a9ce2beb0bdafc0495a60f74ef89aa220ddd1142e27adb23e

Edge TPU Python API link

This API is deprecated: Instead try the Coral Python API.

macOS 10.15 3.5
Windows 10 3.5
Generic Linux 3.X

Source code link

To enable the advancement of edge ML technology and allow for expanded platform compatibility, we're proud to share the source code for the following Coral components:

You might also be interested in the TensorFlow source code.