# LocalAI v4.4.0 — Release & Upgrade Intelligence
- **Product:** LocalAI (https://localai.io)
- **Version:** v4.4.0
- **Released:** Jun 10, 2026
- **Impact Score:** 5/10
- **Breaking Changes:** Not verified
- **Canonical Analysis:** https://changelogs.us/localai/updates/v4.4.0/

## Executive Summary (TL;DR)
LocalAI v4.4.0. Source highlight: 🎉 LocalAI 4.4.0 Release! 🚀

## New Features
- feat: prefix-cache-aware routing for distributed mode
- feat(ds4): layer-split distributed inference
- feat(crispasr): add CrispASR backend — multi-architecture ASR + TTS
- feat(worker): add LOCALAI_PREFETCH_MODELS for boot-time gallery prefetch
- feat(distributed): resumable file uploads via HTTP Content-Range
- feat(localvqe/audio): v1.3 release and add spectrograms to audio transform UI

## Bug Fixes
- **Configurable WebRTC ICE candidates (#10231).** New `LOCALAI_WEBRTC_NAT_1TO1_IPS` / `LOCALAI_WEBRTC_ICE_INTERFACES` knobs fix `/v1/realtime` calls dropping a few seconds in under Docker host networking (unroutable `docker0`/`veth` candidates).
- **Backend capability registration fixes** so the right backend is picked for the right job: register 5 backends missing from `BackendCapabilities` (#10107), and add face/speaker-recognition constants registering `insightface` + `speaker-recognition` (#10110).
- fix(config): register 5 backends missing from BackendCapabilities
- fix(config): register parakeet-cpp as a transcript backend (#9718)
- fix(parakeet-cpp): cublas/hipblas/vulkan builds were silently CPU-only
- fix(nemo): pin texterrors to 1.1.6 for GLIBCXX compatibility