NVIDIA-NeMo/Speech
Some data may be outdated — see provenance labels.
A scalable generative AI framework built for researchers and developers working on Large Language Models, Multimodal, and Speech AI (Automatic Speech Recognition and Text-to-Speech) observed · 2026-09-29
Health v2 · maintenance only
95/100
- Activity 97
- Release rhythm 89
- Longevity 100
How is this computed?
round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10) — computed 2026-10-21. Adoption (stars, forks) is never an input.
- gap_med: 19
- age_days: 2633
- days_rel: 75
- days_push: 22
- n_releases_24m: 24
Adoption not part of the score
18521 stars · 3620 forks observed · 2026-09-29
What it is AI-extracted, prompt v1, taxonomy v1, 2026-10-19, confidence 0.99
NVIDIA NeMo Speech is an open‑source Python toolkit for building, training, and deploying speech AI models such as automatic speech recognition, text‑to‑speech, and speaker‑related tasks. It provides modular neural components, pretrained checkpoints, and scalable multi‑GPU training pipelines.
Use cases
- transcribe audio recordings to text with high accuracy
- generate natural‑sounding speech from text in multiple languages
- identify and separate speakers in multi‑speaker recordings
- fine‑tune speech models on custom datasets
- run streaming ASR or TTS inference in production services
When to choose
- you need a production‑ready, GPU‑accelerated speech AI stack
- you want pretrained models and easy fine‑tuning for ASR, TTS, or speaker tasks
- you require multi‑node, mixed‑precision training for large speech models
When to avoid
- your project is limited to CPU‑only environments with strict latency constraints
- you prefer a lightweight pure‑Python library without PyTorch dependencies
- you need a non‑Python runtime such as Java or JavaScript
Facets
library · maturity active
audio-processing nlp deep-learning llm-inference artificial-intelligence machine-learning deep-learning speech-processing python windows cloud asr tts speaker-diarization speaker-recognition multimodal gpu-accelerated pytorch huggingface natural-language-processing linux macos docker
10 sources
- readme: https://github.com/NVIDIA-NeMo/Speech · fetched 2026-09-29 · 7542fca65f19
- homepage: https://docs.nvidia.com/nemo/speech/nightly/index.html · fetched 2026-10-05 · 88940b7bbfb3
- site_page: https://docs.nvidia.com/nemo/speech/nightly/starthere/install.html · fetched 2026-10-05 · 01e754bc554d
- site_page: https://docs.nvidia.com/nemo/speech/nightly/features/parallelisms.html · fetched 2026-10-05 · ffe16f442542
- site_page: https://docs.nvidia.com/nemo/speech/nightly/features/mixed_precision.html · fetched 2026-10-05 · a27ec8c772f6
- site_page: https://docs.nvidia.com/nemo/speech/nightly/asr/ssl/resources.html · fetched 2026-10-05 · b8ad5cdef481
- site_page: https://docs.nvidia.com/nemo/speech/nightly/asr/speech_classification/resources.html · fetched 2026-10-05 · 42abd7fde7af
- site_page: https://docs.nvidia.com/nemo/speech/nightly/tts/intro.html · fetched 2026-10-05 · 3338cc790bf9
- site_page: https://docs.nvidia.com/nemo/speech/nightly/asr/speaker_diarization/resources.html · fetched 2026-10-05 · b5e462cc79f8
- site_page: https://docs.nvidia.com/nemo/speech/nightly/asr/speaker_recognition/resources.html · fetched 2026-10-05 · fe734b54483c
Member repositories
| Repository | Role | Health v2 |
|---|---|---|
| NVIDIA-NeMo/Speech | main | 95 |
For agents
markdown · JSON · MCP: product_card(name="NVIDIA-NeMo/Speech")
Data as of 2026-10-19T08:05:51.212177+00:00 · Report a problem