nowfound

Life & fun · June 28, 2018

DB

DeepSpeech based automated transcription service

We have been building a DeepSpeech model with our data for the past year and we have recently hit 95% accuracy on the LibriSpeech dataset. That puts us close to the published results for DeepSpeech 2. However our dataset is conversational audio and we do much better with our own internal dataset compared to PaddlePaddle. Here's a blog post on the method we followed to build our models. https://scribie.com/blog/2018/03/continual-learning-for-spee... We have been using this internally in our service and it saves a ton of time and effort during the typing stage. It…

DeepSpeech based automated transcription service launched on June 28, 2018 with 67 votes, #615 of 2,317 launches that month and more than 90% of that year's launches.

In plain words

This automated transcription service uses a DeepSpeech model trained on conversational audio to achieve 95% accuracy on the LibriSpeech dataset. Built over a year of development, it performs better on conversational speech than existing alternatives. The tool is designed for users who need to transcribe audio quickly, though it acknowledges that human transcribers still achieve higher accuracy. The service offers free automated transcripts for a limited time.

written from the facts on this page · September 2026

From the sources

In the maker’s words, at launch

We have been building a DeepSpeech model with our data for the past year and we have recently hit 95% accuracy on the LibriSpeech dataset. That puts us close to the published results for DeepSpeech 2. However our dataset is conversational audio and we do much better with our own internal dataset compared to…

  • DeepGram2016 · deepgram.com · ▲819

    Drag and drop speech transcript editor that learns from you

  • Conversational AI 2.0 From ElevenLabs2025 · elevenlabs.io · ▲652

    Luke from ElevenLabs here. We just released Scribe Realtime v2, our new real-time Speech to Text model. We rebuilt it from the ground up around a new architecture that uses predictive transcription – it anticipates the next likely word before it’s spoken. That makes it both fast (around 150 ms latency) and highly accurate. Scribe v2 Realtime outperforms every low latency transcription model across 30 commonly used EU and Asian languages – 93.5% accuracy, compared to Gemini 2.5 Flash (91.4%), GPT-4o MiniTranscribe (90.7%), and Deepgram Nova 3 (85%). It supports 90+ languages including…

  • AssemblyAI2017 · assemblyai.com · ▲996

    APIs to automatically transcribe and understand audio

  • Moonshine Open-Weights STT modelsFeb 2026 · github.com · ▲316

    Moonshine offers open-weights speech-to-text models that achieve lower word-error rates than OpenAI's Whisper Large v3, built by a six-person startup.

  • Transcriptal2023 · pohonemas33.io · ▲118

    Free AI powered Youtube transcription platform

  • Speechtext.ai2020 · speechtext.ai · ▲12

    AI software for speech to text conversion and audio/video transcription. Get accurate results using domain-specific speech recognition engine! Start for free.

More life & fun this month

the category →
  • H
    HackerBooks▲1,868

    Browse thousands of books recommended in Hacker News comments, 2006–2026. Ranked by how many different people recommend them.

    Life & fun · 21d ago · hackerbooks.org

  • Meet Subvocal. Silent text input for your computer, at the speed of speech. Built by The Subvocal Company.

    Life & fun · 20d ago · subvocal.company

  • A PHP security challenge. Find the hidden bitcoin key. First person to get the wallet wins.

    Life & fun · 29d ago · stealmybitcoin.com

  • Enter any place on Earth. See every satellite that passed over it this week, and the actual pictures they took.

    Life & fun · 13d ago · watched.earth

  • Hacker Times is a newspaper-style publication that curates and reformats stories from Hacker News with original headlines and sources, updated hourly.

    Life & fun · 20d ago · hackertimes.ca

  • TitleArt▲409

    Every classic WordArt style, shape and 3-D effect, in your browser.

    Life & fun · 28d ago · titleart.dvy.io

Is DeepSpeech based automated transcription service yours? Feature it on the home page: $19, or free with the badge →

Source: Show HN post ↗ · launch text quoted in part

Launched alongside, June 2018

the whole month →
  • iOS 12▲705

    Making your device faster, responsive and delightful 📱

    Dev tools · 2018 · apple.com

  • Find the best tools to start a side project, without code 🛠

    Dev tools · 2018

  • Free logos for open source projects

    Dev tools · 2018 · openlogos.org

  • Discover and listen to the world's podcasts on Android.

    Work · 2018 · play.google.com

  • A new way to connect, get more done, and work more fluidly.

    Dev tools · 2018 · apple.com

  • This quick quiz helps you understand why you

    Work · 2018 · whydoiprocrastinate.com