🎯Skills5
Transcribes audio using Sarvam's Saarika v2.5 speech-to-text model with base64 WebSocket streaming, part of the Sarvam AI skill pack built for Indian-language-first AI applications.
Converts text to natural speech using Sarvam AI's Bulbul v3 model, supporting 11 Indian languages with 30+ voice options. Provides REST, HTTP streaming, and WebSocket interfaces with Python and JavaScript/TypeScript SDKs for low-latency audio generation.
A translation skill powered by Sarvam AI that translates text between English and 22+ Indian languages using the Sarvam-Translate and Mayura models. Supports formal style, script selection, speaker gender, and numeral format options for content translation and app localization.
A skill for building conversational voice agents using Sarvam AI with LiveKit or Pipecat frameworks. Integrates Sarvam STT (Saaras v3), TTS (Bulbul v3), and LLM (Sarvam-30B) for real-time voice applications optimized for Indian languages.
A Sarvam AI skill for chat completions using Indian language LLMs (Sarvam-105B and Sarvam-30B) with up to 128K context. It provides an OpenAI-compatible API for building chatbots, Q&A systems, and agents targeting Indian users, with support for streaming and multilingual conversations.