# LocalAI v4.5.6 — Release & Upgrade Intelligence
- **Product:** LocalAI (https://localai.io)
- **Version:** v4.5.6
- **Released:** Jun 30, 2026
- **Impact Score:** 5/10
- **Breaking Changes:** Not verified
- **Canonical Analysis:** https://changelogs.us/localai/updates/v4.5.6/

## Executive Summary (TL;DR)
LocalAI v4.5.6. Source highlight: chore(deps): bump actions/cache from 4 to 6 by @dependabot[bot] in https://github.com/mudler/LocalAI/pull/10593

## New Features
- feat(distributed): SyncedMap component + migrate finetune/quant/agent-tasks to cross-replica state
- feat(distributed): add LOCALAI_DISTRIBUTED_SHARED_MODELS to skip staging on shared volumes (#10556)
- feat(backends): add voice-detect + face-detect ggml backends (replace Python insightface/speaker-recognition)
- feat(realtime): Semantic VAD EOU token
- feat(config): default swa_full:true for sliding-window-attention models

## Bug Fixes
- fix(distributed): missing agent NATS permission
- fix(config): fall back to DefaultContextSize for unparseable GGUFs; pin NVFP4 gallery context_size
- fix(gallery): match mmproj/model quant as a whole token so F16 no longer selects BF16 (#10559)
- fix(distributed): return empty backend list for agent nodes instead of failing backend.list (#10545)
- fix(openresponses): bound resume-stream buffer and enforce response ownership
- fix(ik-llama): port multimodal path to mtmd API and bump to f96eaddb (#10534)