Open Speech and Decision Models Arrive for Apple Devices
What is Introducing AnySpeechRecognizer and AnyDecisionModel: Open Models for Speech and Decisions on Apple Platforms?
AnySpeechRecognizer lets iOS apps run local speech-to-text using models like NVIDIA Parakeet or Alibaba Qwen3-ASR, plus speaker diarization. AnyDecisionModel provides on-device yes/no, multiple-choice and rating answers with probability scores, using MLX, Core AI or llama.cpp.
Who it is for
iOS developers can add accurate, offline speech and decision-making to their apps without paying for cloud APIs.