Nyílt beszédfelismerő és döntési eszköztárat adott ki a Hugging Face

Language models can do a lot. They write, summarize, translate, and call tools. But they're only a small part of what machine learning can do. Say a customer calls your support line. First, you want to know what they said (speech recognition). Then, you want to decide where to send them (decision). You could hand both to a language model, but there are models built for each that are better, faster, and less expensive. In our AnyLanguageModel 1.0 recap, we looked at a year of local and hosted models in Swift. With AnyLanguageModel and swift-huggingface now on their 1.x releases, we're extending the same approach to speech and decisions: familiar Swift APIs with a choice of model. Both packages can download open models from the Hugging Face Hub and run them on device. Computers have been transcribing speech for decades, and in the past few years, they've gotten very good at it. Apple platforms already have an API for it. macOS 26 and iOS 26 introduced SpeechAnalyzer, Apple's new speech-to-text API. You add modules such as SpeechTranscriber and SpeechDetector to an analyzer, give it a file or a stream of audio buffers, and read results from async sequences as they arrive. AssetInventory installs the models, so your app decides when downloads happen. It's a well-designed API, but the only engine behind it is Apple's. AnySpeechRecognizer keeps the API and lets you choose the engine: The package mirrors SpeechAnalyzer, SpeechTranscriber, DictationTranscriber, SpeechDetector, and AssetInventory, with Apple's Speech framework as the default engine. If your code uses that shared API, changing the import is enough. To use Parakeet instead, change only how you create the transcriber: The steps for installing assets, running the analyzer, and reading results stay the same. The README has a complete example and lists the differences from Apple's framework. Parakeet v3 is NVIDIA's Parakeet TDT 0.6B v3. It's a polyglot: it recognizes 25 European languages and works out which one it's hearing, so you don't pass a locale. Parakeet Ultra is Moondream's post-trained take on v3. It was more accurate in most of our measurements. Parakeet Unified is NVIDIA's English-only Parakeet. All three Parakeets run on the package's own Core ML runtime, which supports file and live input, volatile results, word timing, and word confidence. Qwen3-ASR is Alibaba's speech model, in 0.6B and 1.7B sizes. It recognizes 30 languages and 22 Chinese dialects, including Mandarin, Cantonese, Japanese, and Korean, which Parakeet doesn't cover. It detects the language automatically, or you can set one: Qwen3-ASR gives one final result for each segment of up to 40 seconds, without volatile results, word timing, or confidence. It runs on the GPU and is slower than Parakeet, so it's better suited to transcribing files than live input. To add another engine, implement SpeechRecognitionModel for its configuration and SpeechRecognitionExecutor for one module's session. The README includes a small provider that reports placeholder text to help you get started. To run these models in apps, we converted them to Core ML and published them on the Hub: Each repository records its source checkpoint and a checksum for every file, and the package pins it to a fixed revision. The Parakeet repositories add an 8-second encoder for live input, which cuts live analysis time by about a third at the cost of a somewhat higher word error rate. The repository also has the conversion workflows, most of which follow FluidInference's mobius scripts, and checks each conversion against the original model. Downloads happen only when your app asks for them: about 600 MB for each Parakeet, plus about 590 MB for the live encoder, 0.9 to 2.4 GB for Qwen3-ASR, and about 250 MB for Sortformer. Files go in the standard Hugging Face cache, so a Mac app outside the sandbox shares them with Python and other Hugging Face tools. When you transcribe a conversation, you often want to know who said w