---
title: "AlphaSignal Audio: Speech To Text News"
url: "https://alphasignal.ai/topic/audio/speech-to-text"
description: "Latest AI and machine learning news about Speech To Text (Audio) on AlphaSignal."
---

# AlphaSignal Audio: Speech To Text News

Latest AI and machine learning news about Speech To Text (Audio) on AlphaSignal.

## Articles

- [Cactus Compute's Whistle Squeezes 7-Language Speech Recognition Into 16.9 MB](https://alphasignal.ai/news/cactus-compute-s-whistle-squeezes-7-language-speech-recognition-into-16-9-mb.md) (2026-10-09)
- [Open-Source Mimi Brings Live Subtitles to Every App on Your Desktop](https://alphasignal.ai/news/open-source-mimi-brings-live-subtitles-to-every-app-on-your-desktop.md) (2026-10-06)
- [Microsoft's MAI-Transcribe-2-Streaming Tops 38 Models With 2.5% Error Rate](https://alphasignal.ai/news/microsoft-s-mai-transcribe-2-streaming-tops-38-models-with-2-5-error-rate.md) (2026-10-01)
- [Fermion Research's Phonon-2 Beats Whisper at 5.21% WER in Just 164 MB](https://alphasignal.ai/news/fermion-research-s-phonon-2-beats-whisper-at-5-21-wer-in-just-164-mb.md) (2026-10-01)
- [ElevenLabs' Scribe v2 Now Edits Transcripts Without a Separate AI Call](https://alphasignal.ai/news/elevenlabs-scribe-v2-now-edits-transcripts-without-a-separate-ai-call.md) (2026-09-30)
- [Xenova's Whisper Tiny Brings Free Speech Recognition to Any Browser](https://alphasignal.ai/news/xenova-s-whisper-tiny-brings-free-speech-recognition-to-any-browser.md) (2026-09-30)
- [SYSTRAN's Faster Whisper Medium Transcribes 99 Languages 2.4x Faster](https://alphasignal.ai/news/systran-s-faster-whisper-medium-transcribes-99-languages-2-4x-faster.md) (2026-09-29)
- [Fish Audio's transcribe-1-pro Adds Speaker ID and Emotions for $0.36 an Hour](https://alphasignal.ai/news/fish-audio-s-transcribe-1-pro-adds-speaker-id-and-emotions-for-0-36-an-hour.md) (2026-09-28)
- [Norwegian Developer Builds a Raspberry Pi Frame That Draws Birds it Hears](https://alphasignal.ai/news/norwegian-developer-builds-a-raspberry-pi-frame-that-draws-birds-it-hears.md) (2026-09-26)
- [Moondream Shrinks Parakeet Redux Speech Recognition to 178 MB](https://alphasignal.ai/news/moondream-shrinks-parakeet-redux-speech-recognition-to-178-mb.md) (2026-09-25)
- [Sarvam Releases Saaras V4 to Transcribe All 22 Indian Languages at Once](https://alphasignal.ai/news/sarvam-releases-saaras-v4-to-transcribe-all-22-indian-languages-at-once.md) (2026-09-25)
- [Edge0 Ships Audio8 ASR Infinite to Transcribe Speech for Hours Straight](https://alphasignal.ai/news/edge0-ships-audio8-asr-infinite-to-transcribe-speech-for-hours-straight.md) (2026-09-25)
- [NVIDIA's Nemotron 3 Beats 12 Rivals With a 14.72% Speaker Error Rate](https://alphasignal.ai/news/nvidia-s-nemotron-3-beats-12-rivals-with-a-14-72-speaker-error-rate.md) (2026-09-23)
- [Alibaba's Qwen-Audio 3.1 Slashes Voice API Prices by up to 95%](https://alphasignal.ai/news/alibaba-s-qwen-audio-3-1-slashes-voice-api-prices-by-up-to-95.md) (2026-09-23)
- [StepFun's StepAudio 3 ASR Tops Speech Benchmark With 1.7% Error Rate](https://alphasignal.ai/news/stepfun-s-stepaudio-3-asr-tops-speech-benchmark-with-1-7-error-rate.md) (2026-09-22)
- [NetEase Youdao's Confucius4-R2T2 Streams Speech Without Ever Revising Committed Words](https://alphasignal.ai/news/netease-youdao-s-confucius4-r2t2-streams-speech-without-ever-revising-committed.md) (2026-09-22)
- [xAI Ships Grok Voice Transcribe 2.0 With Half the Errors at Same Price](https://alphasignal.ai/news/xai-ships-grok-voice-transcribe-2-0-with-half-the-errors-at-same-price.md) (2026-09-18)
- [LM Studio 1.1.3 Brings Private On-Device Voice Transcription to Linux](https://alphasignal.ai/news/lm-studio-1-1-3-brings-private-on-device-voice-transcription-to-linux.md) (2026-09-15)
- [OpenAI's GPT-Live-1 Cuts Voice Agent Code by 80% at $0.05 a Minute](https://alphasignal.ai/news/openai-s-gpt-live-1-cuts-voice-agent-code-by-80-at-0-05-a-minute.md) (2026-09-10)
- [Tencent's AuK Replaces 16 Speech Tools With One Open-Source Model](https://alphasignal.ai/news/tencent-s-auk-replaces-16-speech-tools-with-one-open-source-model.md) (2026-09-10)
- [Microsoft's MAI-Transcribe-2 Beats ElevenLabs on Accuracy at 54% Lower Price](https://alphasignal.ai/news/microsoft-s-mai-transcribe-2-beats-elevenlabs-on-accuracy-at-54-lower-price.md) (2026-09-03)
- [audio.cpp Runs Speech and Music AI up to 5x Faster Without Python](https://alphasignal.ai/news/audio-cpp-runs-speech-and-music-ai-up-to-5x-faster-without-python.md) (2026-09-01)
- [Meta's Muse Voice Transcribe Beats Every Rival With 3.1% Error Rate](https://alphasignal.ai/news/meta-s-muse-voice-transcribe-beats-every-rival-with-3-1-error-rate.md) (2026-09-01)
- [Google's Gemini 3.5 Transcribe Hits 2.6% Error Rate and Reads Your Screen](https://alphasignal.ai/news/google-s-gemini-3-5-transcribe-hits-2-6-error-rate-and-reads-your-screen.md) (2026-08-26)
- [Kizuna AI Lab's Sokuji Translates Live Calls Offline Without a GPU](https://alphasignal.ai/news/kizuna-ai-lab-s-sokuji-translates-live-calls-offline-without-a-gpu.md) (2026-08-21)
- [Sarvam's Saaras v3 Powers Equal AI's 285M Monthly Indian Calls](https://alphasignal.ai/news/sarvam-s-saaras-v3-powers-equal-ai-s-285m-monthly-indian-calls.md) (2026-08-18)
- [Google DeepMind's SL2T Doubles Sign Language AI Accuracy, Giving Deaf Users a Voice](https://alphasignal.ai/news/google-deepmind-s-sl2t-doubles-sign-language-ai-accuracy-giving-deaf-users-a.md) (2026-08-12)
- [Sarvam AI's Indic DiarBench Tests Speech Recognition Across all 22 Indian Languages](https://alphasignal.ai/news/sarvam-ai-s-indic-diarbench-tests-speech-recognition-across-all-22-indian.md) (2026-08-11)
- [Sarvam AI Cuts Transcription Errors 20% for India's Wealth Manager Dezerv](https://alphasignal.ai/news/sarvam-ai-cuts-transcription-errors-20-for-india-s-wealth-manager-dezerv.md) (2026-08-10)
- [Why your keyboard is the new bottleneck for AI agents (and how to solve it)](https://alphasignal.ai/news/why-your-keyboard-is-the-new-bottleneck-for-ai-agents-and-how-to-solve-it.md) (2026-08-06)
- [Tencent's Hy ASR 3.0 Understands What You Meant, Not Just What You Said](https://alphasignal.ai/news/tencent-s-hy-asr-3-0-understands-what-you-meant-not-just-what-you-said.md) (2026-08-04)
- [Alibaba's Qwen-Audio-3.0-ASR-Flash Hits 95% Accuracy on Medical Jargon](https://alphasignal.ai/news/alibaba-s-qwen-audio-3-0-asr-flash-hits-95-accuracy-on-medical-jargon.md) (2026-07-31)
- [Sarvam AI's Saaras V4 Beats Global English Benchmarks With Single-Pass Multi-Speaker Recognition](https://alphasignal.ai/news/sarvam-ai-s-saaras-v4-beats-global-english-benchmarks-with-single-pass-multi.md) (2026-07-30)
- [MLX-Audio Brings 646-Language Voice Cloning to Apple Silicon Without the Cloud](https://alphasignal.ai/news/mlx-audio-brings-646-language-voice-cloning-to-apple-silicon-without-the-cloud.md) (2026-07-29)
- [xAI's Grok Voice Think Fast 2.0 Hits Sub-Second Response, Beating GPT-Realtime](https://alphasignal.ai/news/xai-s-grok-voice-think-fast-2-0-hits-sub-second-response-beating-gpt-realtime.md) (2026-07-29)
- [Alibaba's Qwen Audio 3.0 Dethrones OpenAI With a 99.2% Reasoning Score](https://alphasignal.ai/news/alibaba-s-qwen-audio-3-0-dethrones-openai-with-a-99-2-reasoning-score.md) (2026-07-28)
- [OpenAI Replaces Whisper With gpt-transcribe, Slashing Errors by 52%](https://alphasignal.ai/news/openai-replaces-whisper-with-gpt-transcribe-slashing-errors-by-52.md) (2026-07-28)
- [Hume AI's VoiceEQ Benchmark Exposes What 1M Human Ratings Reveal About Voice Models](https://alphasignal.ai/news/hume-ai-s-voiceeq-benchmark-exposes-what-1m-human-ratings-reveal-about-voice.md) (2026-07-15)
- [audio.cpp Ships 60+ Audio Models in GGUF Running 5x Faster Than Python](https://alphasignal.ai/news/audio-cpp-ships-60-audio-models-in-gguf-running-5x-faster-than-python.md) (2026-07-14)
- [ByteDance's video-SALMONN 2 Beats GPT-4o by Teaching AI to Hear Videos](https://alphasignal.ai/news/bytedance-s-video-salmonn-2-beats-gpt-4o-by-teaching-ai-to-hear-videos.md) (2026-07-13)
- [Edge0 Releases Audio8-ASR-0.1B to Run Speech Recognition on Any Phone](https://alphasignal.ai/news/edge0-releases-audio8-asr-0-1b-to-run-speech-recognition-on-any-phone.md) (2026-07-10)
- [Moonshot AI's Kimi-Audio Beats GPT-4o on Emotion Control Trained on 13M Hours](https://alphasignal.ai/news/moonshot-ai-s-kimi-audio-beats-gpt-4o-on-emotion-control-trained-on-13m-hours.md) (2026-07-10)
- [Cohere's Open-Source Transcribe Arabic Beats Whisper in 96% of Tests](https://alphasignal.ai/news/cohere-s-open-source-transcribe-arabic-beats-whisper-in-96-of-tests.md) (2026-07-09)
- [Cohere's Cohere Transcribe Arabic Beats Meta's 7B Model at 2B Parameters](https://alphasignal.ai/news/cohere-s-cohere-transcribe-arabic-beats-meta-s-7b-model-at-2b-parameters.md) (2026-07-07)
- [Google DeepMind's SURF Separates Audio Without Ever Hearing Clean Sounds](https://alphasignal.ai/news/google-deepmind-s-surf-separates-audio-without-ever-hearing-clean-sounds.md) (2026-07-07)
- [AssemblyAI's Universal-3.5 Pro Gives Voice Agents Memory Between Turns](https://alphasignal.ai/news/assemblyai-s-universal-3-5-pro-gives-voice-agents-memory-between-turns.md) (2026-07-06)
- [Hugging Face Ships an Open-Source Voice Agent That Ditches OpenAI's Realtime API](https://alphasignal.ai/news/hugging-face-ships-an-open-source-voice-agent-that-ditches-openai-s-realtime-api.md) (2026-07-03)
- [Alibaba's Qwen3-ASR Beats Whisper Across 52 Languages With a Single Open Model](https://alphasignal.ai/news/alibaba-s-qwen3-asr-beats-whisper-across-52-languages-with-a-single-open-model.md) (2026-06-26)
- [Alibaba's Qwen3-ASR Beats GPT-4o Transcribe at a Fraction of the Size](https://alphasignal.ai/news/alibaba-s-qwen3-asr-beats-gpt-4o-transcribe-at-a-fraction-of-the-size.md) (2026-06-26)
- [CJ Pais Built Handy After Breaking a Finger, Now It Has 21K Stars](https://alphasignal.ai/news/cj-pais-built-handy-after-breaking-a-finger-now-it-has-21k-stars.md) (2026-06-24)