Search results

Qwen3-TTS generation is per-step-overhead-bound, not bandwidth-bound — real-time streaming not achievable on M4 Pro with any model size

repo/mlx-audio-swift

Status: Open.

#224 In Blaizzy/mlx-audio-swift;

· raucodes opened on Jul 14, 2026

Support Stable Audio 3 medium

repo/mlx-audio-swift

Status: Open.

#210 In Blaizzy/mlx-audio-swift;

· JacobLinCool opened on Jun 16, 2026

Support MiMo-V2.5-ASR

repo/mlx-audio-swift

Status: Open.

#183 In Blaizzy/mlx-audio-swift;

· ailuntx opened on May 12, 2026

Support translate task for ASR models like Qwen3-ASR

repo/mlx-audio-swift

Status: Open.

#173 In Blaizzy/mlx-audio-swift;

· sysulq opened on Apr 28, 2026

Parakeet mel preprocessing bug in v0.1.2 causes ~3× WER degradation on conversational speech

repo/mlx-audio-swift

Status: Open.

#138 In Blaizzy/mlx-audio-swift;

· ianandrews opened on Mar 30, 2026

Reduce TTS model boilerplate: ~1,000 lines of duplicated loading infrastructure

repo/mlx-audio-swift

Status: Open.

#120 In Blaizzy/mlx-audio-swift;

· beshkenadze opened on Mar 23, 2026

Proposal: MLXAudioKaldiAlignment

repo/mlx-audio-swift

Status: Open.

#85 In Blaizzy/mlx-audio-swift;

· beshkenadze opened on Mar 8, 2026