NVIDIA Riva
GPU-optimized speech AI SDK: FastPitch + HiFi-GAN TTS with custom voices for cloud, on-prem and edge.
About
NVIDIA Riva is the company's GPU-optimized speech AI SDK, generally available since March 2022, covering both speech recognition and text-to-speech for enterprise deployments. The TTS side is built on NVIDIA's NeMo stack — FastPitch acoustic models with HiFi-GAN vocoding — and ships with male and female voices plus Riva Custom Voice, which builds a brand voice from about 30 minutes of recordings. It deploys as containers on NGC for cloud, on-prem, edge and embedded use, scaling to hundreds of thousands of streams.
There is no public per-character pricing: small-scale research and development is free on NGC, and production licensing runs through NVIDIA AI Enterprise (contact sales). NVIDIA claims 2x better accuracy over generic services for the Riva 2.0 generation — a provider claim with no independent public benchmark I could find.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
No independent published benchmarks; NVIDIA's '2x better accuracy' claim is provider marketing
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 8/10
Scored 8/10 (TIER verified, metrics-v1). Evidence: NVIDIA GA announcement (GTC, March 22, 2022); Riva 2.0 release notes; FastPitch + HiFi-GAN stack documented in GTC materials; NGC distribution. Strengths: GPU-optimized throughput, custom voice from 30 minutes of audio, deploy-anywhere containers, enterprise support. Docked for: (1) no public pricing — enterprise sales process; (2) no independent benchmarks; (3) requires NVIDIA GPU infrastructure to shine. License: proprietary (NVIDIA AI Enterprise; free for small-scale R&D).
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 8.0/10
- Maintenance · recency, releases, issue load
- 8.0/10
- Openness · license + docs
- 4.0/10
- Manual review · sanity read + red-flag check
- 8.0/10
- · GA March 22, 2022 (GTC); GPU-optimized speech AI SDK (ASR + TTS)
- · TTS: FastPitch acoustic models + HiFi-GAN vocoding (NeMo stack); Riva Custom Voice from ~30 min audio
- · No public per-unit pricing; deploy cloud/on-prem/edge/embedded via NGC containers
- · NVIDIA logo staged