{"name":"FluidAudio","repo":"FluidInference/FluidAudio","url":"https://midorreal.com/project/fluidinference-fluidaudio","repository":"https://github.com/FluidInference/FluidAudio","description":"Frontier CoreML audio models in your apps — text-to-speech, speech-to-text, voice activity detection, and speaker diarization. In Swift, powered by SOTA open source.","facts":{"license":{"value":"Apache-2.0","source":"https://github.com/FluidInference/FluidAudio","checked":"22 Nov 2026"},"language":{"value":"Swift","source":"https://github.com/FluidInference/FluidAudio","checked":"22 Nov 2026"},"homepage":{"value":"https://docs.fluidinference.com/introduction","source":"https://docs.fluidinference.com/introduction","checked":"22 Nov 2026"},"last_release":{"value":"v0.18.4, 19 Nov 2026","source":"https://github.com/FluidInference/FluidAudio/releases","checked":"22 Nov 2026"},"releases":{"value":"78","source":"https://github.com/FluidInference/FluidAudio/releases","checked":"20 Nov 2026"},"contributors":{"value":"135","source":"https://github.com/FluidInference/FluidAudio/graphs/contributors","checked":"20 Nov 2026"},"archived":{"value":"no","source":"https://github.com/FluidInference/FluidAudio","checked":"22 Nov 2026"}},"stars":3320.0,"writeup":{"what_it_is":"FluidAudio is a Swift SDK for running audio AI models locally on Apple devices, with inference on the Apple Neural Engine. It includes speech recognition, text-to-speech, voice activity detection, and speaker diarization, using open-source CoreML models.","audience":"Swift developers building audio features into macOS and iOS apps","claims":[{"kind":"specific","claim":"Runs inference on the Apple Neural Engine, avoiding GPU and MPS","excerpt":"inference offloaded to the Apple Neural Engine (ANE), resulting in less memory and generally faster inference","status":"not_checked"},{"kind":"specific","claim":"Models are open-source and permissively licensed","excerpt":"All models are publicly available on HuggingFace","status":"not_checked"},{"kind":"specific","claim":"Parakeet TDT v3 supports 25 European languages and Japanese","excerpt":"supporting 25 European languages and Japanese","status":"not_checked"},{"kind":"specific","claim":"Kokoro TTS supports 9 languages","excerpt":"across 9 languages (EN, ES, FR, HI, IT, JA, PT, ZH)","status":"not_checked"},{"kind":"broad","claim":"Integration needs only a few lines of code","excerpt":"that can be integrated with just a few lines of code","status":"too_broad"}],"alternatives":[],"written_by":"AI, from the project's README","date":"2026-10-29"},"why_now":null,"owner_supplied":null,"score":{"verdict":"sleeper","hype":55,"reality":87,"gap":-32,"momentum":0.5,"confidence":60,"version":"score-2.0","rule":"verdict-2.0","calculated_at":"2026-11-22T08:00:00+00:00","verdict_source":"rule"},"early_read":null,"changes":[{"date":"2026-11-19T21:44:37+00:00","text":"Latest release v0.18.4","source":"https://github.com/FluidInference/FluidAudio/releases"},{"date":"2025-06-27T21:55:59+00:00","text":"First release","source":"https://github.com/FluidInference/FluidAudio/releases"},{"date":"2025-06-21T21:09:30+00:00","text":"Repository created","source":"https://github.com/FluidInference/FluidAudio"}],"cite":"FluidAudio is a GitHub project at https://github.com/FluidInference/FluidAudio written mainly in Swift, described there as \"Frontier CoreML audio models in your apps - text-to-speech, speech-to-text, voice activity detection, and speaker diarization. In Swift, powered by SOTA open source\".","attribution":"Data from Mid or Real, https://midorreal.com/project/fluidinference-fluidaudio, checked 22 November 2026. Attribution with a link is required.","still_real":null}