# Hugging Face Transformers v5.2.0 — Release & Upgrade Intelligence
- **Product:** Hugging Face Transformers (https://huggingface.co)
- **Version:** v5.2.0
- **Released:** Feb 16, 2026
- **Impact Score:** 7/10
- **Breaking Changes:** Source flagged
- **Canonical Analysis:** https://changelogs.us/transformers/updates/v5.2.0/

## Executive Summary (TL;DR)
Hugging Face Transformers v5.2.0. Source highlight: <img width="1920" height="1080" alt="image" src="https://github.com/user-attachments/assets/80e37670-6d70-402b-8c8e-ccfb8c32df2d" />

## Breaking Changes & Deprecations
- Breaking changes * :rotating_light: [`Attn`] New attn mask interface everywhere (#42848)
- :rotating_light: This one is quite breaking for super super super old modles: :rotating_light: :rotating_light:

## New Features
- Add Voxtral Realtime (#43769)
- Add GlmMoeDsa (#43858)
- Add VibeVoice Acoustic Tokenizer (#43400)
- Add Voxtral Realtime (#43769)
- Add VibeVoice Acoustic Tokenizer (#43400)

## Improvements
- Improve use of torch.is_autocast_enabled (#43930)

## Bug Fixes
- fix: Prevent AutoTokenizer type mismatch from directory name substrin… (#43791)
- [docs] deploying (#43241)
- [Trainer] Move NEFTune impl to standalone functions (#43714)
- Fix `convert_rope_params_to_dict` so it uses `rope_theta` from the config (#43766)
- Bump dev version (#43777)
- Improved `AGENTS.md` (#43763)

## Upgrade Command
```bash
python -m pip install 'transformers==5.2.0'
```