Speech to text models
built by you.
Cloud based speech to text service. Process in real-time or batch mode. We customize our models to generate the best results for your content.
Turning voice data into text for accessibility, discovery and analysis. Our mission is to make the spoken word accessible to everybody using speech-to-text. With our speech recognition platform, you can turn any audio file into text. The generated text output can then be used to improve digital accessibility by adding subtitles to videos. We see that digital accessibility and greater inclusion is becoming ‘the new standard’ worldwide. For example, as described in the EU Directive 2016/2102, governments must take measures to ensure that everyone has equal access to information. In other parts of the world, we see this trend as well, both required by official rules and coming from inclusive organisations that are driven to make their content accessible to everyone. Do you want to know more about the technology behind speech-to-text?
Customized speech to text models, that work for you. Available in 13 languages but also on demand. Do you need a specific language model that is not yet supported? Contact us! We can build any language!
- Arabic
- Danish
- Dutch
- English
- Filipino
- Finnish
- Flemish
- French
- German
- Italian
- Norwegian
- Spanish
- Swedish
A speech-to-text solution for everyone
Speech-to-text software already has a wide variety of possible applications, and the list is getting bigger all the time . The Media industry, Governments, healthcare, telco and journalism are industries that are already familiar with the benefits of speech-to-text. Check our industry page for in depth case studies.
Everybody can use our speech recognition platform, from single users to enterprises. Whether you are looking to process small amounts of content to subtitle video’s or looking to process multiple channels daily, we have a plan ready to fit your needs.
Individual
- 5 hours p.m.
- Web application
- Subtitle Editor
- Transcript Editor
- Single user
Team
- 24 hours p.m.
- Web application
- Subtitle Editor
- Transcript Editor
- Multiple users
- API Access
Enterprise
- Starts at 10k minutes p.m.
- Cloud or Hosted
- API Access
- Multiple users
- Subtitle Editor
- Transcript Editor
Customize
Build your own model from scratch hosted exclusively for you based on your industry or sector specific content
Custom made speech-to-text models
At Scriptix we build our speech-to-text models in house, which means we can customize them to better fit specific customer needs. All content is created within a different context with its own related audio and vocabulary. Think of the difference between a debate in politics and a conversation between a doctor and a patient. We’ve already built speech-to-text models for different use cases and industries.
What are the key features?
- Confidence scores – retrieve confidence scores on word level
- Timestamps in milliseconds – each word becomes searchable in milliseconds
- Realtime processing – process live streams in all supported languages
- Batch processing – process in batch in all supported languages
- Punctuation – add punctuation to improve readability
- Speaker diarization – label different speakers
- Multichannel processing – process channels separately reduce cross-talk
- Various file formats: retrieve JSON, DOC, VTT, TTML, SRT and SBV formats
Create your own workflow
The Scriptix Speech to Text platform was built with the idea to enable customers to optimize their workflows and products with a safe, stable and scalable speech recognition product. We continuously optimize the Scriptix platform based on requests and wishes from our customers. No need for flashy bells and whistles, the Scriptix Speech to Text platform is meant to be functional and easy-to-integrate. As a result, we now offer a platform that can be integrated in less than a day, enabling our customers and partners to automate all the functionalities we have to offer within their own products and workflows.