All issues

Search results

Qwen fails to function

Status: Open.

#2321 In lmstudio-ai/lmstudio-bug-tracker;

· Zemogiter opened on Aug 24, 2026

Parallel requests invalidate each other's KV cache, causing full context re-prefill on every concurrent request

Status: Open.

#2320 In lmstudio-ai/lmstudio-bug-tracker;

· VladS23 opened on Aug 24, 2026

Loader fails 'failed to allocate buffer for kv cache' for Qwen3.6-35B-A3B and 'compute pp buffers' for GLM-4.6V-Flash; same GGUFs load fine via the bundled llama-server directly (Bionic 1.0.9, runtime 2.29.1, Windows)

Status: Open.

#2319 In lmstudio-ai/lmstudio-bug-tracker;

· acsvalentinacs opened on Aug 23, 2026

MLX models silently ignore requested context length (-c, defaultContextLength, and per-model config all discarded)

Status: Open.

#2318 In lmstudio-ai/lmstudio-bug-tracker;

· livbonito opened on Aug 23, 2026

ARM Linux Oracle Cloud instance install fails

Status: Open.

#2317 In lmstudio-ai/lmstudio-bug-tracker;

· mmmmyah opened on Aug 22, 2026

Recent LM Studio versions no longer support NVIDIA Quadro P4000 (sm_61)

Status: Open.

#2316 In lmstudio-ai/lmstudio-bug-tracker;

· giesalabs opened on Aug 22, 2026

LM Studio Bionic take a lot of times before model start doing someting in chat

Status: Open.

#2315 In lmstudio-ai/lmstudio-bug-tracker;

· avecchi-vsolutions opened on Aug 22, 2026

[Linux] .deb installs LM Studio app icon in an unrecognized hicolor/0x0 directory

Status: Open.

#2314 In lmstudio-ai/lmstudio-bug-tracker;

· Deali-Axy opened on Aug 22, 2026

MCP tools from mcp.json are not attached to chats served via LM Link / Locally

Status: Open.

#2313 In lmstudio-ai/lmstudio-bug-tracker;

· principalwater opened on Aug 22, 2026

[Windows][0.4.21] CUDA 12 runtime partially offloads to system memory despite Max/GPU-only configuration

Status: Open.

#2312 In lmstudio-ai/lmstudio-bug-tracker;

· IanWardell opened on Aug 22, 2026

LM Studio is causing weird lag for my mouse

Status: Open.

#2311 In lmstudio-ai/lmstudio-bug-tracker;

· Artaherzadeh opened on Aug 22, 2026

Incorrect inference on Ryzen AI 395 MAX+with ROCM runtime higher 2.25

Status: Open.

#2310 In lmstudio-ai/lmstudio-bug-tracker;

· dci2002 opened on Aug 21, 2026