First screen, captured .
Looks similar
Show HN: Needle2: 14MB agentic LLM for phones, wearables, smart home and robots(cactuscompute.com)
501 points by HenryNdubuaku | 169 comments
Product page for Cactus Needle 2, an open 14MB agentic LLM for tool calling, device use, and structured extraction on phones, wearables, and tiny devices.
Show HN: Qwen3-4B with the AI writing tics trained out(100toolkit.com)
135 points by erikkannike | 95 comments
Blog post introducing a fine-tuned Qwen3-4B-Instruct model, trained on a MacBook Air, that avoids common AI writing tics and runs locally via Ollama.
VibeVoice: A Frontier Open-Source Text-to-Speech Model(microsoft.github.io)
406 points by lastdong | 150 comments
Project page for Microsoft's VibeVoice, an open-source TTS model generating up to 90 minutes of multi-speaker conversational audio using LLM and diffusion.
Apple's new SpeechAnalyzer API, benchmarked against Whisper and its predecessor(get-inscribe.com)
529 points by get-inscribe | 211 comments
Blog post from Inscribe benchmarking Apple's new SpeechAnalyzer API against its predecessor and OpenAI Whisper on 5,559 test utterances, with transcripts released.
Related stories
Gemini-3.5-Transcribe(blog.google)
338 points by k9294 | 119 comments
Google blog post announcing Gemini 3.5 Transcribe, a speech-to-text model for precise real-time transcription.
Show HN: I trained a 9M speech model to fix my Mandarin tones(simedw.com)
426 points by simedw | 128 comments
Blog post describing a 9M-parameter Conformer-CTC model trained on ~300 hours of speech that grades Mandarin pronunciation and tones in the browser.
Speech Recognition and TTS in less than 500kb(github.com)
544 points by petewarden | 79 comments
GitHub directory for Moonshine's micro build, offering speech recognition and neural text-to-speech small enough to run on microcontrollers like the RP2350.
WhisperSpeech – An open source text-to-speech system built by inverting Whisper(github.com)
405 points by nickmcc | 102 comments
GitHub repository for WhisperSpeech, Collabora's open source text-to-speech system built by inverting OpenAI's Whisper model.
Qwen3.5: Towards Native Multimodal Agents(qwen.ai)
393 points by danielhanchen | 182 comments
Qwen team blog post announcing Qwen3.5, a model family aimed at native multimodal agent capabilities.
Writing a C compiler in 500 lines of Python(vgel.me)
454 points by vgel | 142 comments
Blog post by vgel describing the design and architecture of a small C compiler written in about 500 lines of Python.