# LocalAI v2.23.0 — Release & Upgrade Intelligence
- **Product:** LocalAI (https://localai.io)
- **Version:** v2.23.0
- **Released:** Nov 10, 2024
- **Impact Score:** 7/10
- **Breaking Changes:** Source flagged
- **Canonical Analysis:** https://changelogs.us/localai/updates/v2.23.0/

## Executive Summary (TL;DR)
LocalAI v2.23.0. Source highlight: feat(templates): use a single template for multimodals messages

## Breaking Changes & Deprecations
- Breaking Changes 🛠 * feat(templates): use a single template for multimodals messages by @mudler in https://github.com/mudler/LocalAI/pull/3892 ### Bug fixes :bug:

## New Features
- feat(templates): use a single template for multimodals messages
- feat(vllm): expose 'load_format'
- feat(tts): Implement naive response_format for tts endpoint
- feat(diffusers): allow multiple lora adapters
- feat(ui): move model detailed info to a modal
- feat: allow to disable '/metrics' endpoints for local stats

## Improvements
- chore(p2p): enhance logging

## Bug Fixes
- fix(parler-tts): use latest audiotools
- fix(parler-tts): pin grpcio-tools
- fix(gallery): overrides for parler-tts in the gallery
- fix(grpc): pass by modelpath
- fix(tts): correctly pass backend config when generating model options
- fix(phi3-vision): add multimodal template