# LocalAI v4.0.0 — Release & Upgrade Intelligence
- **Product:** LocalAI (https://localai.io)
- **Version:** v4.0.0
- **Released:** Mar 14, 2026
- **Impact Score:** 7/10
- **Breaking Changes:** Source flagged
- **Canonical Analysis:** https://changelogs.us/localai/updates/v4.0.0/

## Executive Summary (TL;DR)
LocalAI v4.0.0. Source highlight: 🎉 LocalAI 4.0.0 Release! 🚀

## Breaking Changes & Deprecations
- Breaking Changes 🛠 * Remove HuggingFace backend support by @localai-bot in https://github.com/mudler/LocalAI/pull/8971 * chore: drop AIO images by @mudler in https://github.com/mudler/LocalAI/pull/9004
- fix: whisper breaking on cuda-13 (use absolute path for CUDA directory detection) by @localai-bot in https://github.com/mudler/LocalAI/pull/8678

## New Features
- **Disable Option:** Add `LOCALAI_DISABLE_MCP` to completely disable MCP support for security.
- **MLX Distributed (Experimental):** We've added an experimental backend for running distributed workloads using Apple's MLX framework! Check out the docs here.
- **New Audio Backends:** Introduced **fish-speech**, **ace-step.cpp**, and **faster-qwen3-tts** (CUDA-only).
- **WeRTC Realtime:** WebRTC support added to the Realtime API and Talk page for better low-latency audio handling.
- **TTS Improvements:** Added `sample_rate` support via post-processing and multi-voice support for Qwen TTS.
- feat(traces): Use accordian instead of pop-ups

## Improvements
- **Traces:** Improved trace display using accordions for better readability.
- chore(ui): improve errors and reporting during model installation

## Bug Fixes
- **Video Generation:** Fixed model selection dropdown sync and added `vllm-omni` backend detection.
- **Logging:** Fixed watchdog spamming logs when no interval was configured; downgraded health check logs to debug.
- **CUDA Detection:** Improved GPU vendor checks to prevent false CUDA detection on CPU-only hosts with runtime libs.
- **Compatibility:** Renamed `json_verbose` to `verbose_json` for OpenAI spec compliance (fixes Nextcloud integration).
- **Embedding:** Fixed embedding dimension truncation to return full native dimensions.
- **Permissions:** Changed model install file permissions to 0644 to ensure server readability.