Model Catalogue

A catalogue of pre-trained models

  • A fast and efficient deployment experience in just a few clicks
  • A solution based on trusted partners
  • Expertise in speech recognition and word processing
catalogue de modèles OVHcloud
Lettria

Lettria

With Lettria, the structuring of textual data becomes simple, collaborative and effective.

3 pre-trained models are available.

  • Sentiment analysis
    Lettria can automatically classify text data as positive, negative or neutral, depending on the sentiment expressed in the text, using only AI-based algorithms.
     
  • Emotion analysis
    The Lettria API can identify and extract emotion at a sentence level and, using AI-based algorithms, classify it according to the emotion conveyed.
     
  • Entity recognition (coming soon)
    The Lettria API can detect entities in text and automatically classify and label text based on the type of entity mentioned, such as a person, place, or organisation. To achieve this, our algorithms have been trained on large amounts of annotated textual data, manually labelled based on the entities they contain. We use a combination of rules-based algorithms and machine learning, as well as custom solutions that you can create on the Lettria platform, to accurately detect entities in text.

Gladia (coming soon)

Gladia is a solution that offers easy-to-use APIs for processing unstructured data. With its plug-and-play feature, you can integrate the API into your existing systems and start using it immediately.

  • Transcription service
    It includes transcription of audio files in multiple formats, as well as diarisation (separation of speakers) and word-level timestamps.
     
  • Additionalfeatures
    - conversation categorisation into one of 698 subjects (Topic Classification)
    - sentiment detection (positive, negative, neutral)
    - detection of 27 emotions
    - conversation classification for moderation through 19 tags (violence, nsfw, drugs, etc.)
    - translation into 99 languages
Gladia
voxist

Voxist (coming soon)

From voices to knowledge, Voxist provides access to speech recognition technologies thanks to their APIs using the latest innovations. Harness their expertise to transform voices into valuable data and understand your customers, employees and contacts.

Cutting-edge AI models

Thanks to the latest advances in unsupervised learning, our end-to-end models are performing better than some of today's global solutions. Our APIs are both faster and more accurate than existing APIs.

  • Voice recognition
    • Real-time transcription
    • Asynchronous transcription