optimum-intel

library2.0.0pypypiunverified

Optimum Intel extends the Hugging Face Transformers and Diffusers libraries, providing a framework to integrate Intel's specialized tools and libraries like OpenVINO, Neural Compressor, and Intel Extension for PyTorch. It enables optimization, conversion (e.g., to OpenVINO IR format), and accelerated inference of deep learning models on Intel architectures. The library is actively maintained with frequent minor version releases, currently at 1.27.0.

ai-ml

pip install --upgrade-strategy eager "optimum-intel[openvino]"

package ↗homepage ↗

INSTALL

IMPORT

SIG · OPTIMUM-INTEL

optimum-intel

ai-mlpythonv2.0.0

Install

75.8s avg

Import

20603ms

Disk

5530MB

Pass rate

1/ 10

Env Coverage1 / 10

glibc

3.9–3.13

musl

3.9–3.13

Install & Compatibility

Where this runs

tested against v1.27.0 · 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

1/4 runs

py 3.11

✕ build_error

1/4 runs

py 3.12

✕ build_error

1/4 runs

py 3.13

✕ build_error

✓ 75.85s

py 3.9

✕ build_error

✕ timeout

5530MB installed

● package 5530MB

Code

Verified usage

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

OVModelForCausalLM

✓ from optimum.intel import OVModelForCausalLM

✗ from transformers import AutoModelForCausalLM

Use `OVModelForCausalLM` for OpenVINO-optimized causal language models.

OVModelForSeq2SeqLM

✓ from optimum.intel import OVModelForSeq2SeqLM

✗ from transformers import AutoModelForSeq2SeqLM

Use `OVModelForSeq2SeqLM` for OpenVINO-optimized sequence-to-sequence models.

OVStableDiffusionPipeline

✓ from optimum.intel import OVStableDiffusionPipeline

✗ from diffusers import StableDiffusionPipeline

Use `OVStableDiffusionPipeline` for OpenVINO-optimized Diffusers Stable Diffusion pipelines.

INCModelForSequenceClassification

✓ from optimum.intel import INCModelForSequenceClassification

✗ from optimum.intel.lpot.quantization import LpotQuantizerForSequenceClassification

Older versions used `lpot` which was renamed to `neural_compressor`, then simplified to direct import under `optimum.intel`.

This quickstart demonstrates loading a pre-trained sentiment analysis model, converting it to OpenVINO Intermediate Representation (IR) format on the fly using `export=True`, and running inference with a Hugging Face pipeline. Ensure `optimum-intel[openvino]` and `transformers` are installed.

from transformers import AutoTokenizer, pipeline from optimum.intel import OVModelForSequenceClassification model_id = "distilbert-base-uncased-finetuned-sst-2-english" tokenizer = AutoTokenizer.from_pretrained(model_id) # Load and convert the model to OpenVINO IR format on the fly model = OVModelForSequenceClassification.from_pretrained(model_id, export=True) # Run inference classifier = pipeline("text-classification", model=model, tokenizer=tokenizer) results = classifier("Optimum Intel is great!") print(results)

optimum-cli --version

Debug

Known issues

deprecatedThe installation extras for specific backends (e.g., `[openvino]`, `[nncf]`, `[neural-compressor]`, `[ipex]`) via `pip install optimum-intel[...]` are deprecated and will be removed in a future release. Users are encouraged to install `optimum` and its specific backend extras directly or install `optimum-intel` base and then the backend libraries separately.

fix

For future compatibility, install `optimum` with relevant extras (e.g., `pip install optimum[openvino]`) or install `optimum-intel` then individual dependencies like `openvino-dev`.

affects: >=1.27.0

breakingThe `nf4_fp8` quantization modes have been removed. Code relying on these specific quantization modes will break.

fix

Review your quantization configurations and migrate to supported quantization modes, such as INT8 or INT4 weight-only quantization with NNCF.

affects: >=1.27.0

gotchaWhen using OpenVINO Runtime with PyTorch for post-processing (e.g., beam search), OpenVINO's default threading (oneTBB) can interact poorly with PyTorch's OpenMP, leading to performance degradation or delays.

fix

It is recommended to limit the number of CPU threads used by PyTorch with `torch.set_num_threads()` to mitigate this interaction.

affects: All versions

Upgrade

Version history

2.0.0latest on PyPI · released Jun 10, 2026

Audit

Dependencies

optimumrequiredCore Optimum library for hardware optimizations.

transformersrequiredHugging Face Transformers models are the primary target for optimization.

torchrequiredPyTorch backend often used for original models and post-processing.

optimum-onnxrequiredRequired by core `optimum-intel` for ONNX capabilities.

openvinooptionalRequired for OpenVINO runtime and model conversion.

nncfoptionalRequired for Neural Network Compression Framework (NNCF) quantization features.

intel-extension-for-pytorchoptionalRequired for IPEX optimizations.

diffusersoptionalRequired for optimizing and inferencing Hugging Face Diffusers models (e.g., Stable Diffusion).

Agent activity

4 hits · last 30 days

node

Resources

packagehttps://pypi.org/project/optimum-intel/ ↗

homepagehttps://www.intel.com ↗