you can just do things

nvidia is goated parakeet-tdt-0.6b-v3: Multilingual Speech-to-Text Model huggingface.co/nvidia/parakee…

anyone else seeing chatgpt taking ~3s to load? happens pretty consistently and its really annoying since it just hangs > 980 Mbps download > 0 ms: Navigate to chatgpt.com - skeleton load normally > +472 ms: Segment analytics /assets/dz9i23gwxgkzmpfz.js

Senko is now 3x faster on Mac thanks to CoreML. It now processes 1 hour of audio in ~8 seconds on an M3 MacBook, making it by far the the fastest open-src speaker diarization pipeline on Mac, all while maintaining good accuracy.

VoiceInk for iOS is now live on App Store.

New tooling for profiling performance-critical services with Swift -- introducing the Swift Profile Recorder. More here: swift.org/blog/swift-pro…

it’s unfortunate that the gpt-oss talk didn’t even mention mlx - even tho all the demos were on macs then they pulled out the dgx spark preview and it all made sense apple’s marketing team should push their hardware more

Voice on mobile with interruption using Apple’s SpeechAnalyzer, MLX Swift for running the LLM and the following CoreML models: Kokoro, Silero and SegmentAnyText. All running in orchestration. Awni Hannun Fluid Inference

Introducing Coral NPU: A full-stack, open-source platform co-designed by Google Research and Google DeepMind. It provides tools for developers to build the next generation of always-on AI on low-power edge devices. Learn how you can start building today: goo.gle/494bBIs

The future of AI may fit in your hand. Fluid Inference is bringing AI from the cloud to your devices with an open-source platform for local transformer optimization. CEO Brandon Weng (brandon) joins #IntelOnAI to talk cost, speed, and equity. Listen now. Apple:

Today we’re releasing a full suite of models and a dataset: 1️⃣ Omnilingual ASR: A suite of ASR models ranging from 300M to 7B parameters, supporting 1600+ languages 2️⃣ Omnilingual w2v 2.0: a 7B-parameter multilingual speech representation model that can be leveraged for other

after a month of effort, i finally got the eddy-audio repo released for beta, an oss C++ AI audio toolkit optimized for Intel NPU via OpenVINO currently supporting parakeet v2/v3 github.com/FluidInference…

#iOSdev tool alert! 🚨 FluidAudio by is a high-performance Swift framework for on-device speaker diarization and audio processing, designed to meet research standards. github.com/FluidInference…

we have officially crossed 900 stars on fluidaudio, with this current rate of growth, we can expect 1000 before 2026! github.com/FluidInference…

can we get a spending leaderboard

This is the DeepSeek moment for Voice AI. Today we’re releasing Chatterbox Turbo — our state-of-the-art MIT licensed voice model that beats ElevenLabs Turbo and Cartesia Sonic 3! We’re finally removing the trade-offs that have held voice AI back. Fast models sound robotic.

Meta announced a new SAM Audio Model for audio editing that can isolate and extract any sound. Quite clean 🔥

we're excited to support [NVIDIA Sortformer](huggingface.co/nvidia/diar_st…) in FluidAudio. Sortformer handles overlapping speakers and outperforms Pyannote in most use cases especially in noisy environments. Get started here: github.com/FluidInference… Credits to Benjamin Lee for the