Skip to content

LocalAI

v4.2.4 Feature

This release adds 2 notable features for engineering teams evaluating rollout.

Published 21d Model Serving & MLOps
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agents ai api audio-generation decentralized distributed
+12 more
image-generation libp2p llama llm mamba mcp musicgen object-detection rerank stable-diffusion text-generation tts

ReleasePort's take

Moderate signal
editorial:auto 13d

Version v4.2.4 corrects X‑Forwarded-Prefix handling so proxied requests are stripped properly.

Why it matters: If your deployment uses a reverse proxy that strips request prefixes, upgrade to v4.2.4 immediately to avoid routing errors.

Summary

AI summary

Fixed HTTP X-Forwarded-Prefix handling to correctly strip proxy prefixes.

Changes in this release

Feature Medium

parse VRAM budget/usage from vulkaninfo

parse VRAM budget/usage from vulkaninfo

Source: llm_adapter@2026-05-21

Confidence: low

 —
Feature Medium

Add Liquid Audio s2s model and assistant mode on talk page

Add Liquid Audio s2s model and assistant mode on talk page

Source: llm_adapter@2026-05-21

Confidence: low

 —
Dependency Medium

Update ggml-org/llama.cpp to a9883db8ee021cf16783016a60996d41820b5195

Update ggml-org/llama.cpp to a9883db8ee021cf16783016a60996d41820b5195

Source: llm_adapter@2026-05-21

Confidence: low

 —
Dependency Medium

Update TheTom/llama-cpp-turboquant to 5aeb2fdbe26cd4c534c6fa15de73cb5749bd0403

Update TheTom/llama-cpp-turboquant to 5aeb2fdbe26cd4c534c6fa15de73cb5749bd0403

Source: llm_adapter@2026-05-21

Confidence: low

 —
Dependency Medium

Update antirez/ds4 to 0cba357ca1bc0e7510421cc26888e420ea942123

Update antirez/ds4 to 0cba357ca1bc0e7510421cc26888e420ea942123

Source: llm_adapter@2026-05-21

Confidence: low

 —
Dependency Medium

Update ikawrakow/ik_llama.cpp to 949bb8f1d660fc1264c137a6f3dbd619375f6134

Update ikawrakow/ik_llama.cpp to 949bb8f1d660fc1264c137a6f3dbd619375f6134

Source: llm_adapter@2026-05-21

Confidence: low

 —
Dependency Medium

Update ggml-org/whisper.cpp to 3e9b7d0fef3528ee2208da3cdb873a2c53d2ae2f

Update ggml-org/whisper.cpp to 3e9b7d0fef3528ee2208da3cdb873a2c53d2ae2f

Source: llm_adapter@2026-05-21

Confidence: low

 —
Bugfix Medium

cascade-clean stale node_models rows and filter routing by healthy status

cascade-clean stale node_models rows and filter routing by healthy status

Source: llm_adapter@2026-05-21

Confidence: high

 —
Bugfix Medium

honor X-Forwarded-Prefix when proxy strips the prefix

honor X-Forwarded-Prefix when proxy strips the prefix

Source: llm_adapter@2026-05-21

Confidence: high

 —
Bugfix Medium

close truncate-then-read race in agent_jobs.json persistence

close truncate-then-read race in agent_jobs.json persistence

Source: llm_adapter@2026-05-21

Confidence: high

 —
Bugfix Medium

parse OpenAI-spec tool_choice in /v1/chat/completions

parse OpenAI-spec tool_choice in /v1/chat/completions

Source: llm_adapter@2026-05-21

Confidence: high

 —
Other Medium

update docs version mudler/LocalAI

update docs version mudler/LocalAI

Source: llm_adapter@2026-05-21

Confidence: low

 —
Other Medium

publish missing :latest-* and :v<X>-* singleton image tags

publish missing :latest-* and :v<X>-* singleton image tags

Source: llm_adapter@2026-05-21

Confidence: low

 —
Full changelog

What's Changed

Bug fixes :bug:

  • fix(distributed): cascade-clean stale node_models rows + filter routing by healthy status by @localai-bot in https://github.com/mudler/LocalAI/pull/9754
  • fix(http): honor X-Forwarded-Prefix when proxy strips the prefix by @Dennisadira in https://github.com/mudler/LocalAI/pull/9614
  • fix(agentpool): close truncate-then-read race in agent_jobs.json persistence by @localai-bot in https://github.com/mudler/LocalAI/pull/9811
  • fix(middleware): parse OpenAI-spec tool_choice in /v1/chat/completions by @Anai-Guo in https://github.com/mudler/LocalAI/pull/9559

Exciting New Features 🎉

  • feat: also parse VRAM budget/usage from vulkaninfo by @eglia in https://github.com/mudler/LocalAI/pull/9800
  • feat(realtime): Add Liquid Audio s2s model and assistant mode on talk page by @richiejp in https://github.com/mudler/LocalAI/pull/9801

Other Changes

  • chore: :arrow_up: Update ggml-org/llama.cpp to a9883db8ee021cf16783016a60996d41820b5195 by @localai-bot in https://github.com/mudler/LocalAI/pull/9796
  • chore: :arrow_up: Update TheTom/llama-cpp-turboquant to 5aeb2fdbe26cd4c534c6fa15de73cb5749bd0403 by @localai-bot in https://github.com/mudler/LocalAI/pull/9740
  • docs: :arrow_up: update docs version mudler/LocalAI by @localai-bot in https://github.com/mudler/LocalAI/pull/9805
  • chore: :arrow_up: Update antirez/ds4 to 0cba357ca1bc0e7510421cc26888e420ea942123 by @localai-bot in https://github.com/mudler/LocalAI/pull/9806
  • chore: :arrow_up: Update ikawrakow/ik_llama.cpp to 949bb8f1d660fc1264c137a6f3dbd619375f6134 by @localai-bot in https://github.com/mudler/LocalAI/pull/9807
  • chore: :arrow_up: Update ggml-org/whisper.cpp to 3e9b7d0fef3528ee2208da3cdb873a2c53d2ae2f by @localai-bot in https://github.com/mudler/LocalAI/pull/9808
  • ci(image): publish missing :latest-* and :v-* singleton image tags by @localai-bot in https://github.com/mudler/LocalAI/pull/9812

Full Changelog: https://github.com/mudler/LocalAI/compare/v4.2.3...v4.2.4

View diff on GitHub

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Share on X Share on Bluesky

Track LocalAI

Get notified when new releases ship.

Sign up free

About LocalAI

LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.

All releases →

Related context

Beta — feedback welcome: [email protected]