Issues

• VoxCPM Backend Installs CUDA 13 PyTorch Stack which results in noise rather than Speech

  bugunconfirmed
  mudler/LocalAI#12356 · Filter by author shopsD opened on Sep 29, 2026

• Open Responses API doesn't accept "output_text" as an input Item

  bugunconfirmed
  mudler/LocalAI#12337 · Filter by author Zoarial94 opened on Sep 28, 2026
  1

• feat: add automatic MCP transport selection for 2024-11-05 / 2025-03-26 / 2025-06-18 vs 2025-11-25

  enhancement
  mudler/LocalAI#12262 · Filter by author edk4971 opened on Sep 25, 2026
  2

• Voice Design Model Support in Voice/Personality Library

  enhancement
  mudler/LocalAI#12219 · Filter by author shopsD opened on Sep 23, 2026
  1

• Add Pixal3D backend

  enhancement
  mudler/LocalAI#12201 · Filter by author iki77 opened on Sep 22, 2026
  1

• localai cannot run the recommended model

  bugunconfirmed
  mudler/LocalAI#12081 · Filter by author guystreeter opened on Sep 16, 2026
  6

• Socket activation not working

  bugunconfirmed
  mudler/LocalAI#12024 · Filter by author pasqui23 opened on Sep 13, 2026
  1

• Gallery model audio-cpp-sortformer-diarization fails to install: SHA mismatch, reproducible across retries

  mudler/LocalAI#11995 · Filter by author 66656c6c6f7574 opened on Sep 12, 2026
  1

• Agent pool Postgres DSN is logged in cleartext, including the password, while the auth DSN is redacted

  mudler/LocalAI#11991 · Filter by author 66656c6c6f7574 opened on Sep 12, 2026
  1

• vllm-omni backend (cuda13) fails to load: editable install points at build path, and unpinned vllm-omni main no longer matches backend.py

  mudler/LocalAI#11972 · Filter by author alaschgari opened on Sep 11, 2026
  1

• omp gets no response from localai

  bugunconfirmed
  mudler/LocalAI#11928 · Filter by author pasqui23 opened on Sep 8, 2026
  3

• intel-sycl-f16-stablediffusion-ggml: SIGABRT during CGO call

  area/backendsarea/containerbugneed-more-informationos/gnu-Linuxunconfirmedwaiting-from-reporter
  mudler/LocalAI#11896 · Filter by author knightpp opened on Sep 6, 2026
  2

• Model Gallery sizing ("YOUR HOST") ignores registered Distributed Mode nodes

  area/p2parea/webuibugneed-more-informationwaiting-from-reporter
  mudler/LocalAI#11814 · Filter by author AMXELA-Official opened on Sep 1, 2026
  4

• Add Embedding, Reranker (and optional OCR) model fields to Agent configuration for building RAG pipelines at the agent level

  area/apiarea/webuienhancement
  mudler/LocalAI#11753 · Filter by author marcorossi85 opened on Aug 27, 2026
  1

• Upgrade backends fails if auto-upgrade is enabled

  area/backendsbugconfirmedhigh prio
  mudler/LocalAI#11720 · Filter by author toalex77 opened on Aug 25, 2026
  1

• Qwen3.5 llama-cpp tool calling fails when function has "strict": true

  area/apiarea/llama.cppbugconfirmed
  mudler/LocalAI#11709 · Filter by author sat-eng opened on Aug 24, 2026

• RWKV.cpp backend

  area/backendsenhancementmodel/rwkvnew-backend
  mudler/LocalAI#11704 · Filter by author marcorossi85 opened on Aug 24, 2026

• User-controlled Multi-Agent Conversation (Human-in-the-loop)

  area/webuienhancement
  mudler/LocalAI#11703 · Filter by author marcorossi85 opened on Aug 24, 2026
  1

• Agent response disappears when navigating away via "Status" or switching chats in chat section

  bugunconfirmed
  mudler/LocalAI#11702 · Filter by author marcorossi85 opened on Aug 24, 2026
  1

• Instella-MoE models in the LocalAI gallery is not supported by any backends

  area/backendsbugconfirmed
  mudler/LocalAI#11681 · Filter by author areian opened on Aug 23, 2026
  1

• audio-cpp using a fish-audio doesn't work with referance audio/text

  area/backendsbugunconfirmed
  mudler/LocalAI#11626 · Filter by author ALameLlama opened on Aug 20, 2026
  1

• metal-whisper: ggml_metal_library_init fails to compile embedded Metal shader library on macOS 26.5.2 (Apple M4 Pro) — flash-attn kernel's threadgroup half4x4 array rejected

  area/backendsbugconfirmedos/macOSupstream issue
  mudler/LocalAI#11624 · Filter by author kamilsa opened on Aug 20, 2026

• Vision/multimodal silently dropped with use_tokenizer_template on the python backends (sglang, vllm): images reach the engine but the prompt carries no media placeholder

  area/backendsarea/vllmbugconfirmed
  mudler/LocalAI#11621 · Filter by author pos-ei-don opened on Aug 20, 2026
  1

• Voice Library Improvements

  enhancement
  mudler/LocalAI#11617 · Filter by author ALameLlama opened on Aug 20, 2026
  2

• API to list TTS voices

  enhancement
  mudler/LocalAI#11604 · Filter by author pinage404 opened on Aug 19, 2026
  1