---
title: "Intel's AutoRound v0.14 Shrinks 200B+ Models Without a Single GPU"
subtitle: "Intel's AutoRound quantization toolkit gets model-free quantization, stronger MXFP4, diffusion model support, and multi-GPU scaling in v0.14"
url: "https://alphasignal.ai/news/intel-s-autoround-v0-14-shrinks-200b-models-without-a-single-gpu"
author: "AlphaSignal Newsroom"
author_url: "https://alphasignal.ai/editorial-team"
company: "AlphaSignal"
company_url: "https://alphasignal.ai/company/alphasignal"
published: "2026-07-07T12:49:11.000Z"
updated: "2026-07-08T01:11:48.756Z"
type: "Repo"
topics:
  - "Gpus"
  - "Open Source"
subtopics:
  - "Quantization"
  - "Edge Deployment"
upvotes: 13
source_url: "https://github.com/intel/auto-round"
comments: 0
truncated: true
---

# Intel's AutoRound v0.14 Shrinks 200B+ Models Without a Single GPU

> Intel's AutoRound quantization toolkit gets model-free quantization, stronger MXFP4, diffusion model support, and multi-GPU scaling in v0.14

Intel's [AutoRound](https://github.com/intel/auto-round) , a post-training quantization toolkit for LLMs and vision-language models , has a new release on the way. Intel researcher HaihaoShen announced AutoRound v0.14 on X, highlighting four headline additions: model-free quantization, improved MXFP4, WAN/diffusion model support, and multi-card quantization. The release follows a rapid cadence of updates that has seen the project grow to over 1,500 GitHub stars.

AutoRound's core idea is to make quantization smarter than simple rounding. It introduces three trainable parameters per quantized tensor , a rounding offset, and learned clipping range controls , then applies signed gradient descent to jointly optimize rounding and clipping, minimizing block-wise output reconstruction error. The result is that models compressed to 2–4 bits retain far more accuracy than naive approaches.

## What's new in v0.14

The four features called out in the release announcement each address a real friction point in production quantization workflows:

* **Model-free quantization.** AutoRound now provides free devices for calibration-free quantization via pure RTN mode, and `auto-round-rtn`

_Preview only. Full text requires an AlphaSignal Pro account._

## Sources

- [intel/auto-round GitHub repository](https://github.com/intel/auto-round)
- [Releases · intel/auto-round](https://github.com/intel/auto-round/releases)
- [Intel AutoRound Enables Faster & More Efficient Quantized LLM Models — WCCFTech](https://wccftech.com/intel-autoround-faster-more-efficient-llm-models-intel-gpus-cuda-cresent-island-fp8-mxfp8-mxfp4/)
- [Intel Quantization Support — vLLM docs](https://docs.vllm.ai/en/v0.18.0/features/quantization/inc/)
- [Microscaling Quantization — Intel Neural Compressor docs](https://intel.github.io/neural-compressor/latest/docs/source/3x/PT_MXQuant.html)