Conversation AI NIMs

NVIDIA RIVA NIM APIs provide easy access to state-of-art (SOTA) models that are essential for creating enterprise-grade speech and translation applications, providing fast real-time capabilities and exceptional accuracy. Developers can employ these APIs for creating sophisticated Q&A bots, video-conferencing transcription services, translation tools, and AI- driven multilingual contact center assistants that are tailored to specific domains, industries, and customer interactions. RIVA NIM models leverage the NVIDIA software ecosystem, including CUDA, TensorRT, and Triton, enabling immediate GPU acceleration.

RIVA ASR (Automatic Speech Recognition) NIM

Transcribes spoken English with exceptional accuracy.

RIVA TTS (Text-To-Speech) NIM

Synthesizes English speech from text and predicts duration and pitch.

RIVA NMT (Neural Machine Translation) NIM

Translates text from one language to another.

📘

Available Models

You can access the available models here

Select models are available as downloadable container images and supported with an NVIDIA AI Enterprise entitlement. These select models have additional OpenAI API spec details for running self-hosted localized NIMs. Please refer to the Downloadable NIMs section for more details.