# intel/ipex-llm

Accelerate local LLM inference and finetuning (LLaMA, Mistral, ChatGLM, Qwen, DeepSeek, Mixtral, Gemma, Phi, MiniCPM, Qwen-VL, MiniCPM-V, etc.) on Intel XPU (e.g., local PC with iGPU and NPU, discrete GPU such as Arc, Flex and Max); seamlessly integrate with llama.cpp, Ollama, HuggingFace, LangChain, LlamaIndex, vLLM, DeepSpeed, Axolotl, etc.

Repository: https://github.com/intel/ipex-llm
Canonical: https://ross.abutalabs.com/products/ipex-llm
Language: Python
License: Apache-2.0
License Family: permissive
Topics: pytorch, llm, transformers, gpu
Archived: true
Last push: 2026-01-28T17:32:40+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 64, release rhythm 40, longevity 100
- inputs: {"age_days": 3656, "days_push": 217, "days_rel": 511, "gap_med": 1, "n_releases_24m": 2}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 8859, forks 1430 (observed 2026-08-28T04:10:26.068054+00:00)

## What it is
IPEX-LLM is a PyTorch LLM acceleration library for Intel hardware (iGPU, NPU, Arc/Flex/Max GPUs, and CPU), offering low-bit quantization (FP8/FP6/FP4/INT4) and integrations with llama.cpp, Ollama, vLLM, HuggingFace, LangChain, and more. The project has been archived by Intel and no longer receives maintenance, bug fixes, or security updates.

## Use cases
- run local LLM inference on Intel Arc GPU
- run DeepSeek or Qwen models on Intel iGPU or NPU
- finetune LLaMA or Mistral on Intel GPUs
- accelerate Ollama or llama.cpp on Intel hardware
- quantize LLMs to INT4 or FP4 for local inference
- run 671B models on one or two Arc A770 GPUs

## When to choose
- you are locked into Intel XPU hardware and need an existing, tested LLM acceleration path
- you need historical reference for Intel GPU LLM optimization techniques
- you maintain a legacy deployment already built on ipex-llm

## When to avoid
- starting a new project, since the repo is archived with known security issues
- you need ongoing support, bug fixes, or new model support
- you are on NVIDIA/AMD hardware where better-maintained alternatives exist
- security-sensitive or multi-tenant environments

## Facets
- artifact type: library
- maturity: abandoned
- function: llm-inference, llm-training, gpu-computing, machine-learning
- domain: large-language-models, machine-learning, deep-learning, gpu-computing
- platform: python, windows
- tags: intel-xpu, low-bit-quantization, ollama, llama-cpp, vllm, finetuning, archived, linux, gpu

## Member repositories
- intel/ipex-llm (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:26.068054+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:24:46.444149+00:00, confidence not recorded.
  - readme: https://github.com/intel/ipex-llm (fetched 2026-08-28T04:10:26.068054+00:00, sha 9873c75e977d)
- Data as of 2026-08-30T08:39:29.467469+00:00.