Voice Writer

BlogPricingTry NowWrite 5x Faster with AI-Powered Voice WritingSpeak your mind, and let our AI fix your grammarTry it NowsoyeahIwaskindathinkingwemaybeshouldlikecancelthedemofornoworatleastpushitbecausehonestlyIdon'tthinkthethingistotallyreadyandwestillgottafixsomestuffandlikeIhaven'theardbackfromthemanywayright↓Ithinkweshouldpostponethedemofornow.It'snotfullyready,andwestillneedtofixafewissues.Also,wehaven'theardbackfromtheclientyet.⚡Fast & PolishedReal-time transcription and grammar correction🧠Adaptive StyleLearns your writing tone from examples🌐Browser-NativeNo app, no install, works wherever you write🌍Multilingual30+ languages supportedHow it WorksStep 1: Speak Your ThoughtsJust start talking. Our speech engine picks up every word, no matter how you phrase it.Step 2: Our AI Perfects Your GrammarOur grammar engine rewrites your speech into polished, professional sentences - automatically.Step 3: Copy and PasteCopy your clean writing into Gmail, Notion, or wherever you're working.Perfect for Writing:EmailsBlog postsSocial mediaReportsPricingPay as you go, no subscription requiredTry Pro free for 14 days. Keep using basic features forever.FreeFor individuals$0USD / monthBasic AI grammar correctionBrowser's built-in speech recognitionEnglish onlyCreate accountPro14-day free trialFor individuals and teams$10/month$80/year if billed annuallyEverything in Free, plus:Premium speech recognitionSupport for 20+ languagesCustomizable grammar correctionAdjust tone with a clickGet startedVoice Writer for ChromeInstall our browser extension to use Voice Writer on any website.Add to Chrome Voice WriterOthersReal-world accuracy✅ Optimized for accent and noise benchmarks⚠️ Varies widely across conditionsStreaming latency✅ Under 300ms⚠️ Transcribes after recording full phraseGrammar correction LLM✅ Real-time, unified with speech⚠️ Separate or absent📊 View full benchmark resultsData and PrivacyVoice Writer only listens while you're actively speaking, and only when the app is open. The microphone never runs in the background.Your data is stored securely and may be used to improve model accuracy and enhance your experience.We work with trusted third-party providers who follow strict security standards when processing data.TeamBai LiMachine learning engineer, PhD in natural language processingLinkedinElaine YeSoftware engineer (Ex-Amazon)LinkedinTry Voice Writer today!Try Voice Writer→BlogLeaderboardCareersPrivacy© 2026 Efficient NLP. All rights reserved. --- BlogPricingTry NowTry Voice WriterSign upExpanding our Free Plan for everyoneJul 19, 2025We’re excited to announce a major update to our pricing: Voice Writer now offers a forever-free tier with unlimited usage that gives everyone access to our core features with no time limit, no usage caps.Read MoreA Streaming Proxy for Real-Time ASR: Abstracting Speech ProvidersJul 9, 2025How does Voice Writer turn speech into grammatically correct writing in real time? Most voice apps record everything first and only process it afterward. Voice Writer works differently: we use streaming ASR and LLMs to deliver sub-second latency as you speak. In this post, we’ll walk through the real-time architecture that enables low latency, high reliability, and the ability to seamlessly swap out speech providers as better models become available.Read MoreWrite using your voice in 30+ languagesMay 31, 2025Voice Writer now understands you in over 30 languages. Whether you're writing in English, Korean, or Portuguese, you can now speak your thoughts, and we'll turn them into polished, grammatically correct text.Read MoreHow to train your own custom writing toneApr 5, 2025Everyone writes differently—and you shouldn’t have to sound like the same AI as everyone else. With today's update, you now have more control over how your writing sounds. Choose between standard (a clear, polished tone suitable for most professional writing), casual (a more relaxed and conversational tone), or a custom tone that you define, based on your own writing style.Read MoreChoose Your Writing Tone and LLM ProviderMar 8, 2025Not all writing is the same - sometimes, you need a polished, professional tone for an important document, and other times, a casual, friendly style works better. In this update, we are introducing customizable writing tone and AI model selection!Read MoreThe Best Speech Recognition API in 2025: A Head-to-Head ComparisonFeb 6, 2025Speech-to-text technology has improved dramatically in recent years, but with so many options available, which API delivers the best accuracy? In this benchmark, I test and compare the major speech recognition APIs, including cloud-based services from AWS, Google Cloud, and Microsoft Azure, two startups specializing in ASR: Assembly AI and Deepgram, the open-source OpenAI Whisper model, and Google Gemini, a large language model that can process speech input.Read MoreMulti-Document Editing and SyncingDec 15, 2024One long-awaited feature we're excited to roll out is multi-document editing! Voice Writer now allows you to work on unlimited documents at the same time. If you have notes for different projects or just want to keep your ideas organized, you can now use different documents and keep your writing style separate depending on the project.Read MoreVoice Writer can now fix grammar in real timeOct 25, 2024Writing just got easier—Voice Writer now polishes your grammar in real time as you speak, with near-zero latency! No extra steps, no delays—just seamless, polished writing as you dictate. Here’s what it looks like in action:Read MoreThe Architecture of Chrome Extension Permissions: A Deep DiveSep 21, 2024Chrome extensions are incredibly popular tools that enhance your browser's functionality. You probably have a bunch of them installed - these extensions can do things like store passwords, manage tabs, change words on a webpage as you type, block ads, and more and go beyond what the usual website security model allows. Given how much sensitive and personal data your browser stores (browsing history, passwords, and banking credentials, etc) how does a browser ensure the security of extensions? That’s what this post will cover: The Chrome Extension Permission System.Read MoreIntroducing Voice Writer Chrome Extension: Speech-to-text on any websiteAug 30, 2024Excited to announce the launch of the Voice Writer Chrome Extension, now available on the Chrome Web Store! With this extension, you can use Voice Writer’s powerful speech-to-text and grammar correction capabilities directly on any website, making writing faster and easier than ever before.Read MoreA distraction-free AI writing experienceMay 5, 2024Voice Writer has a fresh new look! Taking inspiration from modern text editor designs and user feedback, we’ve redesigned our interface to create a minimalist and distraction-free experience. Here's what it looks like now:Read MoreGrammar correction that keeps your formatting intactApr 11, 2024A quick Voice Writer update: you can now fix grammar while preserving the original formatting of your document! This means your document structure remains untouched, while your writing gets the polish it needs.Read MoreFrom 750 words to 3,000: How to write blog posts faster using speech-to-textApr 7, 2024Recently I've been using Voice Writer every day to write everything from emails and Slack updates to blog posts and other documents. But just how much faster does it make me? That’s what I set out to measure today with a little experiment.Read MoreIntroducing Voice Writer: AI-powered dictation and grammar correctionMar 22, 2024Writing is hard. Sometimes, you have the perfect idea in your head, but getting it into words can feel like an uphill battle. That’s writer’s block—it takes time and effort to transform messy thoughts into polished, coherent writing. And all the time spent fixing grammar and formatting, and the whole process feels like a chore. Read MoreTry Voice Writer today!Try Voice Writer→BlogLeaderboardCareersPrivacy© 2026 Efficient NLP. All rights reserved. --- BlogPricingTry NowTry Voice WriterSign up🏆 Real-World Speech-to-text API LeaderboardThis leaderboard compares the best automatic speech recognition (ASR) models and APIs across real-world conditions like background noise, non-native accents, and technical vocabulary.We built this benchmark to find the best speech recognition API for Voice Writer — an AI writing tool that transcribes speech and corrects grammar and wording in real time. Only real-world results matter here, so this benchmark is fully independent and unbiased.⏱️ Leaderboard last updated: October 2025Category:OverallCleanNoisyAccentedSpecialistEvaluation:?Formatted WERRaw WERInclude streaming?SystemMean WERStd DevPrice (per hour)GPT-4o Transcribe?5.4%4.6%$0.36Gemini 2.5 Pro?5.6%5.2%$0.22Gemini 2.5 Flash?6.7%5.1%$0.14ElevenLabs?6.8%6.0%$0.35AssemblyAI?6.8%4.7%$0.15Whisper Large?7.2%5.1%LocalDeepgram?7.6%4.7%$0.26Speechmatics?7.6%5.5%$0.40AssemblyAI (Streaming)?8.1%5.1%$0.15AWS Transcribe?8.3%5.1%$1.44Deepgram (Streaming)?9.7%5.7%$0.46Whisper Small?9.7%5.9%LocalAWS (Streaming)?9.9%5.1%$1.44Microsoft?10.3%5.4%$0.18Whisper (Streaming)?12.4%16.3%LocalGoogle Speech?13.1%5.8%$0.96Apple Dictation?16.5%6.3%LocalSpeech ConditionsMany existing benchmarks use unrealistic language (e.g., LibriSpeech), very short clips (e.g., Common Voice), or are so widely used for benchmarking that API providers are likely to overfit to them. This makes them less reliable for real-world evaluation, especially for longer audio.To create a fair and relevant comparison for applications like Voice Writer, we built a custom dataset covering four categories:Clean Speech: High-quality TED Talk clips from native English speakers.Noisy Speech: The clean TED clips with added hospital background noise to simulate real-world environments.Accented Speech: Wikipedia readings by speakers with non-native English accents (such as Chinese and Indian accents).Specialist Speech: Technical abstracts from recent machine learning and physics papers, synthesized by a text-to-speech (TTS) system.Each clip is 1–2 minutes long, allowing us to evaluate not only word recognition but also punctuation and capitalization over longer passages.Evaluation MetricsWe measured performance using two types of Word Error Rate (WER):Formatted WER (also called unnormalized WER) — capitalization, punctuation, and formatting are treated as important. This produces higher WER and is the more relevant metric for most writing-related applications.Raw WER (also called normalized WER) — punctuation and capitalization are stripped, and variations are standardized (e.g., "color" vs "colour", "two" vs "2"). This metric is more relevant for machine-processing use cases where formatting is not required.We also distinguish between batch and streaming transcription on the leaderboard:Batch — the full audio file is uploaded and transcribed at once.Streaming — audio is streamed in real time to the provider, and transcripts are generated incrementally. Streaming models typically have lower accuracy than batch models due to limited context.System Inclusion CriteriaTo keep the leaderboard focused, we only include:Major ASR products from large tech companiesStartups with 100+ employees or $100M+ in fundingWidely used open-source modelsAdditional DetailsFor a deeper dive into the benchmark setup, audio samples, and methodology, see this blog post or this video.Note: the blog and video reflect an earlier version of the benchmark from January 2025, so some data, evaluation methods, and models may differ.For questions or inquiries, please contact bai@voicewriter.io.Try Voice Writer today!Try Voice Writer→BlogLeaderboardCareersPrivacy© 2026 Efficient NLP. All rights reserved.