Registry / ai-ml / ethos-u-vela

ethos-u-vela

JSON →
library5.1.0pypypi✓ verified 81d ago

Neural network model compiler for Arm Ethos-U NPUs. Version 5.0.0, released approximately monthly.

pip install ethos-u-vela
INSTALL
IMPORT
SIG · ETHOS-U-VELA
E
ethos-u-vela
ai-mlpythonv5.1.0
Install
3.8s avg
Import
Disk
94MB
Pass rate
4/ 10
Env Coverage4 / 10
glibc
3.93.13
musl
3.93.13
Install & Compatibility
Where this runs
tested against v5.1.0 · pip install
no network on importno background threads
Install × environment matrix
Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux
musl
glibc
py 3.10
✕ build_error
✓ 3.9s
py 3.11
✕ build_error
✓ 3.7s
py 3.12
✕ build_error
✓ 3.8s
py 3.13
✕ build_error
✓ 3.9s
py 3.9
✕ build_error
✕ build_error
94MB installed
● package 94MB
Code
Verified usage

Verified import paths — ran on the pinned version, not inferred.

vela
import ethosu.vela
import vela

After installing ethos-u-vela, use the vela module to compile TFLite models for Ethos-U NPUs.

import vela # Compile a TFLite model vela.main(['--output-dir', './output', 'model.tflite'])
vela --version
Debug
Known issues
breakingThe CLI entry point changed from 'vela' to 'ethos-u-vela' after version 3.x. Running 'vela' directly may fail.
fix
Use 'ethos-u-vela' as the command or ensure the vela script is in PATH.
affects: <4.0.0
deprecatedSupport for TensorFlow 1.x models was removed in v4.0.0. Only TFLite models (TensorFlow 2.x) are accepted.
fix
Convert your model to TFLite format using TensorFlow 2.x.
affects: >=4.0.0
gotchaThe package name is 'ethos-u-vela' but the importable module is 'vela'. This confuses new users.
fix
Use 'import vela' after 'pip install ethos-u-vela'.
affects: all
Upgrade
Version history
5.1.0latest on PyPI · released Jun 4, 2026
Audit
Dependencies

No dependency data recorded yet.

Agent activity
8 hits · last 30 days
node
8
Resources
ethos-u-vela — pip install ethos-u-vela · libregistry