Skip to content

LocalAI

v4.3.5 Feature

This release adds 1 notable feature for engineering teams evaluating rollout.

Published 5d 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

Summary

AI summary

Broad release touches Other Changes, chore, Bug fixes :bug, and Exciting New Features 🎉.

Changes in this release

Feature Medium

Honors per‑request reasoning_effort on chat completions for fine‑grained control.

Honors per‑request reasoning_effort on chat completions for fine‑grained control.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Dependency Low

Updates vllm-project/vllm cu130 wheel to version 0.22.0.

Updates vllm-project/vllm cu130 wheel to version 0.22.0.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Dependency Low

Updates ggml-org/whisper.cpp to commit f24588a272ae8e23280d9c220536437164e6ed28.

Updates ggml-org/whisper.cpp to commit f24588a272ae8e23280d9c220536437164e6ed28.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Dependency Low

Updates mudler/rf-detr.cpp to commit 65c0ffcc9a9bc9dae38252f63d0417c9845a6cf7.

Updates mudler/rf-detr.cpp to commit 65c0ffcc9a9bc9dae38252f63d0417c9845a6cf7.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Dependency Low

Adds one new model via the gallery agent to the model gallery.

Adds one new model via the gallery agent to the model gallery.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Bugfix Medium

Fixes tool-call JSON leaks into content with stream+tools on tokenizer-template models.

Fixes tool-call JSON leaks into content with stream+tools on tokenizer-template models.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Bugfix Medium

Fixes double‑emission of tool‑call when autoparser is active in OpenAI compatibility layer.

Fixes double‑emission of tool‑call when autoparser is active in OpenAI compatibility layer.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Bugfix Medium

Stops backend processes synchronously on shutdown to avoid hanging.

Stops backend processes synchronously on shutdown to avoid hanging.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Bugfix Medium

Validates auto‑detected XML tool‑call names for robustness with glm‑4.5/Hermes.

Validates auto‑detected XML tool‑call names for robustness with glm‑4.5/Hermes.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Bugfix Medium

Tracks intentional stops, preventing clean shutdowns from being misread as crashes.

Tracks intentional stops, preventing clean shutdowns from being misread as crashes.

Source: llm_adapter@2026-05-30

Confidence: high

 —
Full changelog

What's Changed

Bug fixes :bug:

  • fix: tool-call JSON leaks into content with stream+tools on tokenizer-template models (#10052) by @localai-bot in https://github.com/mudler/LocalAI/pull/10057
  • fix(openai): stop streaming tool-call double-emission when autoparser is active by @bozhouDev in https://github.com/mudler/LocalAI/pull/10055
  • fix(application): stop backend processes synchronously on shutdown by @richiejp in https://github.com/mudler/LocalAI/pull/10058
  • fix(functions): validate auto-detected XML tool-call names — robust glm-4.5/Hermes guard (#9722, supersedes #9940) by @localai-bot in https://github.com/mudler/LocalAI/pull/10059
  • fix(model): track intentional stops, stop misreading clean shutdowns as crashes by @richiejp in https://github.com/mudler/LocalAI/pull/10060

Exciting New Features 🎉

  • feat(reasoning): honor per-request reasoning_effort on chat completions by @localai-bot in https://github.com/mudler/LocalAI/pull/10082

Other Changes

  • chore: :arrow_up: Update mudler/rf-detr.cpp to ecf64d7f7f20d73ebd906a983f398ed287256320 by @localai-bot in https://github.com/mudler/LocalAI/pull/10035
  • docs: :arrow_up: update docs version mudler/LocalAI by @localai-bot in https://github.com/mudler/LocalAI/pull/10046
  • chore: :arrow_up: Update antirez/ds4 to 22393e770ea8eb7501d8718d6f66c6374004e03f by @localai-bot in https://github.com/mudler/LocalAI/pull/10047
  • chore: :arrow_up: Update leejet/stable-diffusion.cpp to 0e4ee04488159b81d95a9ffcd983a077fd5dcb77 by @localai-bot in https://github.com/mudler/LocalAI/pull/10048
  • chore: :arrow_up: Update ggml-org/llama.cpp to 751ebd17a58a8a513994509214373bb9e6a3d66c by @localai-bot in https://github.com/mudler/LocalAI/pull/10049
  • chore: :arrow_up: Update ikawrakow/ik_llama.cpp to 6eff055a0cc0e427a6849cfcb5de531b4b82d667 by @localai-bot in https://github.com/mudler/LocalAI/pull/10050
  • chore: :arrow_up: Update ggml-org/whisper.cpp to c932729a304f7d9eb5354afa38624cfa86a780cf by @localai-bot in https://github.com/mudler/LocalAI/pull/10051
  • test(react-ui): cover models gallery empty-state reset flow by @Oceankj in https://github.com/mudler/LocalAI/pull/10019
  • test(utils): cover path verification, sanitization, and unique naming by @TLoE419 in https://github.com/mudler/LocalAI/pull/9978
  • chore(model gallery): :robot: add 1 new models via gallery agent by @localai-bot in https://github.com/mudler/LocalAI/pull/10061
  • chore: :arrow_up: Update ikawrakow/ik_llama.cpp to 8960c5ba5ee9db30ba838304373aa4dbec9f7cbd by @localai-bot in https://github.com/mudler/LocalAI/pull/10077
  • chore: :arrow_up: Update vllm-project/vllm cu130 wheel to 0.22.0 by @localai-bot in https://github.com/mudler/LocalAI/pull/10079
  • chore(model-gallery): :arrow_up: update checksum by @localai-bot in https://github.com/mudler/LocalAI/pull/10081
  • docs: :arrow_up: update docs version mudler/LocalAI by @localai-bot in https://github.com/mudler/LocalAI/pull/10074
  • chore: :arrow_up: Update mudler/rf-detr.cpp to 65c0ffcc9a9bc9dae38252f63d0417c9845a6cf7 by @localai-bot in https://github.com/mudler/LocalAI/pull/10075
  • chore: :arrow_up: Update ggml-org/whisper.cpp to f24588a272ae8e23280d9c220536437164e6ed28 by @localai-bot in https://github.com/mudler/LocalAI/pull/10078

New Contributors

  • @bozhouDev made their first contribution in https://github.com/mudler/LocalAI/pull/10055
  • @Oceankj made their first contribution in https://github.com/mudler/LocalAI/pull/10019
  • @TLoE419 made their first contribution in https://github.com/mudler/LocalAI/pull/9978

Full Changelog: https://github.com/mudler/LocalAI/compare/v4.3.4...v4.3.5

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]