Issues

• submit() has no cancellation path — a non-streaming request whose client gives up holds the worker to max_tokens (non-streaming counterpart of #269)

  cubist38/mlx-openai-server#321 · Filter by author Camekazi opened on Aug 29, 2026

• multimodal: BatchGenerator.__init__() got an unexpected keyword argument 'kv_bits' (dependency skew)

  cubist38/mlx-openai-server#320 · Filter by author tonioloewald opened on Aug 6, 2026

• support mlx-vlm 0.6.x

  cubist38/mlx-openai-server#318 · Filter by author sajozsattila opened on Jul 30, 2026

• Resource limit (499000) exceeded

  cubist38/mlx-openai-server#317 · Filter by author alexmateescu opened on Jun 28, 2026

• Image generation returns 500 (There is no Stream(gpu, 0) in current thread) on mlx >= 0.31.2

  cubist38/mlx-openai-server#314 · Filter by author tokenfires opened on Jun 13, 2026

• Sliding-window models (Gemma 3/4, gpt-oss) fail: "There is no Stream(gpu, N) in current thread"

  cubist38/mlx-openai-server#312 · Filter by author kemp-rv opened on Jun 7, 2026

• Step-3.5 -> Error loading model: 'PreTrainedConfig' object has no attribute 'max_position_embeddings'

  cubist38/mlx-openai-server#308 · Filter by author therealmobasha opened on May 16, 2026

• Handler process crashed upon TimeoutError

  cubist38/mlx-openai-server#307 · Filter by author therealmobasha opened on May 14, 2026

• Any plan on reranker model supporting?

  cubist38/mlx-openai-server#306 · Filter by author tyh-carl opened on May 14, 2026

• [bug] KeyError: '__value__' in prompt cache under concurrent requests

  cubist38/mlx-openai-server#304 · Filter by author doobidoo opened on May 14, 2026

• Support mlx-audio

  cubist38/mlx-openai-server#303 · Filter by author psuckrow-routine opened on May 13, 2026

• BatchScheduler does not honor mid-stream client disconnect (regression of #269 for the v1.8 continuous-batching path)

  cubist38/mlx-openai-server#302 · Filter by author supersteves opened on May 13, 2026

• Feature request: explicit model unload endpoint

  cubist38/mlx-openai-server#300 · Filter by author grumdahl opened on May 8, 2026

• Validation error for ChatCompletionChunk

  cubist38/mlx-openai-server#298 · Filter by author therealmobasha opened on May 4, 2026
  1

• Feat: include custom timeout flag

  cubist38/mlx-openai-server#293 · Filter by author therealmobasha opened on May 2, 2026
  4

• Failed to generate text stream: expected string or bytes-like object, got 'NoneType'

  cubist38/mlx-openai-server#281 · Filter by author therealmobasha opened on Apr 9, 2026

• 5 second example results in immediate crash

  cubist38/mlx-openai-server#275 · Filter by author abra-code opened on Apr 6, 2026

• Single-Model MTP Support?

  cubist38/mlx-openai-server#204 · Filter by author TomLucidor opened on Feb 14, 2026

• Q: Speculative Decoding Support?

  cubist38/mlx-openai-server#177 · Filter by author TomLucidor opened on Jan 31, 2026
  3

• Periodic segfaults: "Command buffer execution failed: Discarded (victim of GPU error/recovery) "

  cubist38/mlx-openai-server#169 · Filter by author jverkoey opened on Jan 28, 2026
  1

• Multiple caching modes

  cubist38/mlx-openai-server#166 · Filter by author blightbow opened on Jan 28, 2026

• Tequila/Sherry + "BitNet" Support

  cubist38/mlx-openai-server#161 · Filter by author TomLucidor opened on Jan 27, 2026

• Claude skills for creating new parsers

  cubist38/mlx-openai-server#145 · Filter by author TomLucidor opened on Jan 15, 2026
  7

• Qwen3-Omni & Global Server spinup

  cubist38/mlx-openai-server#80 · Filter by author DPBattaglia opened on Oct 21, 2025

• FEAT: Parakeet compatability?

  cubist38/mlx-openai-server#77 · Filter by author DPBattaglia opened on Oct 16, 2025
  3