nvidia-speech-nim
Use NVIDIA Speech NIM microservices for speech production and voice workflows, including self-hosted ASR/STT, TTS, text translation, speech-to-speech pipeline design, Riva Python client integration, deployment planning, GPU/runtime sizing, licensing/privacy review, and production QA for transcription, captioning, voice agents, localization, and synthetic voice output.
Profile →npx skills add calesthio/generative-media-skills --skill nvidia-speech-nim --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
The skill guides an agent to plan or implement a speech workflow using NVIDIA Speech NIM microservices as a deployable containerized stack. It delineates separate ASR, TTS, and NMT containers accessed over HTTP, gRPC, or WebSocket APIs, while the agent remains responsible for model/profile choice, orchestration, rights checks, audio custody, and QA.
How it works
- Classify the job into ASR/STT, TTS, NMT, speech-to-speech translation, or full-duplex voice chat.
- Refer to documented capability map showing that ASR, TTS, and NMT are standalone containers, and the application calls NIM APIs.
- Select at deployment time using CONTAINER_ID and NIM_TAGS_SELECTOR, consulting support matrices for profiles, memory, and language support.
- For ASR: choose appropriate streaming or offline profiles, considering language, diarization, and translation requirements; ensure hardware/software boundaries are met (GPU compute capability, VRAM, OS, driver, and Docker versions).
- For TTS: pick voices and synthesis mode (offline/streaming) aligning with latency and localization needs; mind hardware/software boundaries and endpoint limits.
- For NMT: decide when translation is needed; use NMT NIM and model names like riva-translate-1_6b and terminology controls.
- API route selection: HTTP REST, gRPC, or WebSocket with endpoints provided for health, readiness, and realtime sessions; refer to NVIDIA Riva Python Client for ready-to-use scripts.
- Follow production heuristics: preprocess audio, measure WER/CER, tune latency and diarization, manage voice cloning approvals, and validate localization terminology and chunking.
- Adhere to rights, consent, privacy, and licensing steps before generating or transmitting user speech, including custody separation, rights to output voice, and production license paths.
When to use it
Use when the task requires planning or implementing a speech workflow with NVIDIA Speech NIM microservices rather than a generic speech API. Trigger considerations include ASR, TTS, translation, and speech-to-speech translation needs, plus production QA and licensing reviews.
What it can touch
- Tools: claude-code, codex, copilot, cursor
- API surfaces: HTTP REST, gRPC, WebSocket, and Riva Python Client scripts for ready-made calls
- Containers and deployment profiles controlled via CONTAINER_ID and NIM_TAGS_SELECTOR
Caveats
- Documentation notes that NIM containers expose health, version, metadata, model, and Prometheus-like metrics; readiness checks are advised prior to production audio.
- Rights and licensing steps emphasize consent, custody separation, and production license paths (NVIDIA AI Enterprise or production-ready partner endpoints), with explicit references to terms and terms-of-service sources.
- The skill references several NVIDIA sources and requires adherence to model-specific notes and hardware boundaries; no guarantees are stated beyond these documented requirements.
# NVIDIA Speech NIM Use this skill when an agent must plan or implement a speech workflow with NVIDIA Speech NIM microservices rather than a generic speech API. Treat NVIDIA Speech NIM as a deployable containerized speech stack: the application talks to independent ASR, TTS, and NMT NIM containers over HTTP, gRPC, or WebSocket APIs, and the agent remains responsible for model/profile choice, orche
What does the nvidia-speech-nim skill do?
Use NVIDIA Speech NIM microservices for speech production and voice workflows, including self-hosted ASR/STT, TTS, text translation, speech-to-speech pipeline design, Riva Python client integration, deployment planning, GPU/runtime sizing, licensing/privacy review, and production QA for transcription, captioning, voice agents, localization, and synthetic voice output.
How do I install it?
Run `npx skills add calesthio/generative-media-skills --skill nvidia-speech-nim --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From calesthio/generative-media-skills, a repository with 112 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.