# NVIDIA Magpie TTS Review

> NVIDIA Magpie TTS review: A small, GPU-efficient 9-language TTS checkpoint for teams already in the NVIDIA NeMo/Riva ecosystem; commercially usable open.

![NVIDIA Magpie TTS logo](https://www.versusref.com/logos/nvidia-magpie.png)

A small, GPU-efficient 9-language TTS checkpoint for teams already in the NVIDIA NeMo/Riva ecosystem; commercially usable open weights, but zero-shot voice cloning was removed from the open release and it caps generations at about 20 seconds.

Among the 46 text-to-speech tools we track, NVIDIA Magpie TTS has the 26th-widest language coverage.

See pricing · Open source · facts verified Jul 20, 2026

## What we know about NVIDIA Magpie TTS

This is our verified profile of NVIDIA Magpie TTS, a text-to-speech apis platform - a small, GPU-efficient 9-language TTS checkpoint for teams already in the NVIDIA NeMo/Riva ecosystem; commercially usable open weights, but zero-shot voice cloning was removed from the open release and it caps generations at about 20 seconds. Every fact about NVIDIA Magpie TTS below carries the source it came from and the day we checked it.

NVIDIA Magpie TTS does not publish a public per-minute rate we have been able to verify, so the price figure here stays blank until we can confirm one. We would rather show nothing than a guessed number; if you have a current NVIDIA Magpie TTS quote, it is the fastest way for us to close that gap.

On capabilities, NVIDIA Magpie TTS covers self-host / on-prem option, and does not offer instant voice cloning. Each of those is verified against NVIDIA Magpie TTS's own docs or dashboard, not marketing copy.

Among the 46 text-to-speech tools in our matrix, NVIDIA Magpie TTS leads with the 26th-widest language coverage; the fact sheet below has the raw numbers behind that placement.

In total we track 11 verified facts for NVIDIA Magpie TTS today, each linking the primary source it came from so you can check our work - and vendor claims we have not measured ourselves are labeled as such on the NVIDIA Magpie TTS fact sheet below.

## Fact sheet

### Capabilities

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Instant voice cloning | ✗  No | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |
| Languages supported | 9 languages | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |

### Compliance & trust

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Self-host / on-prem option | ✓  Yes | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |
| Model weights license | NVIDIA Open Model License Agreement | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |

### Build experience

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Official SDKs | NVIDIA NeMo Framework (Python) for local inference | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |
| Output formats | 22.05 kHz audio (neural codec output) | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |
| Max input per request | Up to 20s speech per request | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |

### Commercial

| Fact | Value | Verified | Source |
| --- | --- | --- | --- |
| Model size (parameters) | 357M | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |
| Hardware to self-host | NVIDIA A10, A30, A100, or H100 GPU; Linux preferred | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |
| Hosted API available | Yes | Jul 20 | [source](https://build.nvidia.com/nvidia/magpie-tts-multilingual/modelcard) |
| Project maintenance status | Active | Jul 20 | [source](https://huggingface.co/nvidia/magpie_tts_multilingual_357m) |

Source: https://www.versusref.com/tts/tools/nvidia-magpie/