--- title: WhisperSpeech: Exploring new horizons in text-to-speech tech description: Collabora has introduced WhisperSpeech, an open-source text-to-speech model built on properly licensed datasets and unrestricted code, making it suitable for... canonical: https://daily.dev/posts/whisperspeech-exploring-new-horizons-in-text-to-speech-tech-y0imjwepu twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: WhisperSpeech: Exploring new horizons in text-to-speech tech | daily.dev og:description: Collabora has introduced WhisperSpeech, an open-source text-to-speech model built on properly licensed datasets and unrestricted code, making it suitable for... og:url: https://daily.dev/posts/whisperspeech-exploring-new-horizons-in-text-to-speech-tech-y0imjwepu og:image: https://api.daily.dev/og/posts/Y0imJwePU.png og:image:alt: WhisperSpeech: Exploring new horizons in text-to-speech tech og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # WhisperSpeech: Exploring new horizons in text-to-speech tech **[Collabora Blog](https://daily.dev/sources/collabora)** · 5 min read · 0 upvotes · 0 comments ## Summary Collabora has introduced WhisperSpeech, an open-source text-to-speech model built on properly licensed datasets and unrestricted code, making it suitable for commercial use. The model reverses the operation of OpenAI's Whisper speech recognition model to synthesize natural-sounding speech from text. It uses Meta's Encodec audio codec and the Vocos vocoder, dividing synthesis into two phases: Text-to-Semantic (T2S) and Semantic-to-Acoustic (S2A). Potential applications include audio editing, IVR systems, public announcements, audiobooks, and podcasts. Collabora plans to release larger, higher-quality models and is seeking partnerships for integration and deployment. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How does WhisperSpeech generate speech from text at an architectural level? WhisperSpeech reverses the Whisper speech recognition model's operation: it takes text as input, generates embeddings, and transforms them into sound through two phases. First, a Text-to-Semantic (T2S) model produces semantic tokens (500kbps) with phonetic and prosodic attributes. Then a Semantic-to-Acoustic (S2A) model uses Meta's Encodec to compress audio into acoustic tokens (1.5kbps), which the Vocos vocoder converts into final speech output. _Developers building TTS pipelines track open-source model architectures like this on daily.dev._ ### What open-source components does WhisperSpeech rely on? WhisperSpeech builds on three key open-source components: OpenAI's Whisper model (reversed for TTS), Meta's Encodec audio codec for compressing audio into acoustic tokens at 1.5kbps, and the Vocos vocoder from charactr for final audio rendering. The architecture was also inspired by Google Research's SPEAR TTS paper, though Google never released its code or weights. _Teams evaluating open-source speech stacks find component-level breakdowns like this useful on daily.dev._ ## Similar posts on daily.dev - [Cohere Transcribe: state-of-the-art speech recognition](https://daily.dev/posts/cohere-transcribe-state-of-the-art-speech-recognition-bopofp9tb) · Hacker News · 0 upvotes · 0 comments - [Real-Time translated speech pipeline with Whisper and Soprano](https://daily.dev/posts/real-time-translated-speech-pipeline-with-whisper-and-soprano-wqus4v8ef) · DigitalOcean Community · 0 upvotes · 0 comments - [Type with your voice on Linux using this Whisper-based app](https://daily.dev/posts/type-with-your-voice-on-linux-using-this-whisper-based-app-jsd1jlcfz) · omg\! ubuntu\! · 0 upvotes · 0 comments --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#open-source](https://daily.dev/tags/open-source), [#deep-learning](https://daily.dev/tags/deep-learning), [#text-to-speech](https://daily.dev/tags/text-to-speech) [View this post on daily.dev](https://daily.dev/posts/whisperspeech-exploring-new-horizons-in-text-to-speech-tech-y0imjwepu) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"WhisperSpeech: Exploring new horizons in text-to-speech tech","url":"https://daily.dev/posts/whisperspeech-exploring-new-horizons-in-text-to-speech-tech-y0imjwepu","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/whisperspeech-exploring-new-horizons-in-text-to-speech-tech-y0imjwepu"},"datePublished":"2026-08-11T08:56:14.861Z","dateModified":"2026-08-11T09:41:21.050Z","description":"Collabora has introduced WhisperSpeech, an open-source text-to-speech model built on properly licensed datasets and unrestricted code, making it suitable for...","image":"https://media.daily.dev/image/upload/s--HRgLpUt6--/f_auto/v1722860399/public/Placeholder%2003","thumbnailUrl":"https://media.daily.dev/image/upload/s--HRgLpUt6--/f_auto/v1722860399/public/Placeholder%2003","isAccessibleForFree":true,"articleSection":"Collabora Blog","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Collabora Blog","logo":"https://media.daily.dev/image/upload/s--ZGZ0FjqS--/f_auto,q_auto/v1786438306/logos/collabora?_a=BAMAMicg0","url":"https://daily.dev/sources/collabora"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/whisperspeech-exploring-new-horizons-in-text-to-speech-tech-y0imjwepu","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,open-source,deep-learning,text-to-speech","timeRequired":"PT5M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Collabora Blog","item":"https://daily.dev/sources/collabora"},{"@type":"ListItem","position":3,"name":"WhisperSpeech: Exploring new horizons in text-to-speech tech"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/whisperspeech-exploring-new-horizons-in-text-to-speech-tech-y0imjwepu#faq","mainEntity":[{"@type":"Question","name":"How does WhisperSpeech generate speech from text at an architectural level?","acceptedAnswer":{"@type":"Answer","text":"WhisperSpeech reverses the Whisper speech recognition model's operation: it takes text as input, generates embeddings, and transforms them into sound through two phases. First, a Text-to-Semantic (T2S) model produces semantic tokens (500kbps) with phonetic and prosodic attributes. Then a Semantic-to-Acoustic (S2A) model uses Meta's Encodec to compress audio into acoustic tokens (1.5kbps), which the Vocos vocoder converts into final speech output. Developers building TTS pipelines track open-source model architectures like this on daily.dev."}},{"@type":"Question","name":"What open-source components does WhisperSpeech rely on?","acceptedAnswer":{"@type":"Answer","text":"WhisperSpeech builds on three key open-source components: OpenAI's Whisper model (reversed for TTS), Meta's Encodec audio codec for compressing audio into acoustic tokens at 1.5kbps, and the Vocos vocoder from charactr for final audio rendering. The architecture was also inspired by Google Research's SPEAR TTS paper, though Google never released its code or weights. Teams evaluating open-source speech stacks find component-level breakdowns like this useful on daily.dev."}}]} ```