Sarvam AI launches open-source models

The Indian artificial intelligence lab Sarvam AI has announced a new family of models, marking a significant step towards the adoption of open-source AI solutions. The new offering includes large language models, as well as specialized models for text-to-speech, speech-to-text, and document analysis.

Model Details

The model suite includes:

  • Language models with 30 billion parameters.
  • Language models with 105 billion parameters.
  • A text-to-speech model.
  • A speech-to-text model.
  • A vision model designed for document analysis and interpretation.

Sarvam AI's open-source approach could foster greater transparency and collaboration in artificial intelligence development, allowing developers and companies to customize and adapt the models to their specific needs. For those evaluating on-premise deployments, there are trade-offs to consider, and AI-RADAR offers analytical frameworks on /llm-onpremise to evaluate the different options.