Independent project. Not a U.S. government website.

USASI
Model family

NVIDIA Parakeet

Family overview — summarizes releases; licenses and availability belong to each release.

Maintained by NVIDIA (NeMo)51

Parakeet is NVIDIA's family of automatic speech recognition models built on the FastConformer encoder in the NeMo toolkit, released with CTC, RNN-T, TDT, and hybrid TDT-CTC decoders. The first checkpoints were published on Hugging Face in December 2023; the card for one of them, parakeet-rnnt-1.1b, says it was developed jointly by the NVIDIA NeMo and Suno.ai teams. Later releases include the English parakeet-tdt-0.6b-v2 (April 2025), the 25-language parakeet-tdt-0.6b-v3 (August 2025), and parakeet-unified-en-0.6b (April 2026), which handles offline and streaming transcription in one model; there are also language-specific checkpoints such as Japanese, Danish, and Vietnamese.12345Fact reviewed Oct 1, 2026

Last reviewedEntry updated Documented release Dec 2023

Releases assessed

Each release is assessed on its own. This family has no family-wide openness label.
Releases in the NVIDIA Parakeet family
ReleaseWeightsTier (USASI rubric v0.2)LicenseReleased
NVIDIA parakeet-tdt-0.6b-v3Weights: PublicModel-disclosure tier (USASI rubric v0.2): Open-weightCC-BY-4.0, Apache-2.0Aug 14, 2025
NVIDIA parakeet-unified-en-0.6bWeights: PublicModel-disclosure tier (USASI rubric v0.2): Open-weightNVIDIA Open Model License Agreement, Apache-2.0Apr 7, 2026

What it is useful for

Transcribing speech to text in batch or streaming settings, with punctuation, capitalization, and timestamps in recent releases, for uses the cards list such as voice assistants, transcription and captioning services, and voice analytics.34Fact reviewed Oct 1, 2026

Organization context

U.S. eligibility

Project eligibility rests on documented governing or maintaining entities, not on contributors.

Eligible · basis: U.S. headquarters

The Parakeet checkpoints are published in NVIDIA's Hugging Face organization and announced in NVIDIA's NeMo Speech repository, and the cards describe training with NVIDIA's NeMo toolkit. NVIDIA's principal executive offices are in Santa Clara, California, per its Form 10-Q for the quarter ended July 26, 2026. The parakeet-rnnt-1.1b card says that early checkpoint was developed jointly with the Suno.ai team.15326

Assessed Oct 1, 2026

Sources

  1. 1.
    Hugging Face model listing for nvidia Parakeet models (API) (external site: huggingface.co)

    Hugging Face · Other · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026

  2. 2.
    nvidia/parakeet-rnnt-1.1b model card (external site: huggingface.co)

    NVIDIA (Hugging Face) · Model card · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026

  3. 3.
    nvidia/parakeet-tdt-0.6b-v3 model card (external site: huggingface.co)

    NVIDIA (Hugging Face) · Model card · published Aug 14, 2025 · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026

  4. 4.
    nvidia/parakeet-unified-en-0.6b model card (external site: huggingface.co)

    NVIDIA (Hugging Face) · Model card · published Apr 7, 2026 · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026

  5. 5.
    NVIDIA-NeMo/Speech README (external site: raw.githubusercontent.com)

    NVIDIA (GitHub) · Repository · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026

  6. 6.
    NVIDIA Corporation Form 10-Q for the quarter ended July 26, 2026 (external site: sec.gov)

    U.S. Securities and Exchange Commission (EDGAR) · Filing · accessed Oct 1, 2026 · evidence reviewed Oct 1, 2026

Support Us

Help keep USASI useful.

Optional. No USASI account required. Payment takes place on the linked provider’s website (Buy Me a Coffee).

About supporting this project