Model · Voice and speech

NVIDIA Parakeet

A free speech-recognition model from NVIDIA that turns recordings in 25 European languages into punctuated text with timestamps. Commercial use is allowed if you credit NVIDIA. Needs technical setup and a GPU.

Go to NVIDIA Parakeet’s website

At a glance

Open or closed
Fully open
Weights and code downloadable under a standard open licence
Weights under CC BY 4.0, which permits any use with credit.
source · checked 12 Oct 2026 · checked on the maker's own page
Licence
CC BY 4.0 (Permissive (e.g. Apache 2.0, MIT))
source · checked 12 Oct 2026 · checked on the maker's own page
Can I sell what I make?
Yes, with conditions
  • You must credit the maker
Card: 'ready for commercial/non-commercial use'. CC BY 4.0 requires credit for the model; whether this extends to transcripts is not stated.
source · checked 12 Oct 2026 · checked on the maker's own page
Cost
Free option: Free to download; you pay for your own hardware
Plans below.
How you use it
On your own computer
Runs with NVIDIA NeMo (pip install nemo_toolkit[asr]), NeMo-Speech.cpp with a GGUF file, or Transformers. Up to 24 minutes of audio per pass, or 3 hours with local attention.
source · checked 12 Oct 2026 · checked on the maker's own page
Skill needed
Technical
source · checked 12 Oct 2026 · checked, but the page was unclear or partial
Made by
NVIDIA
Model card does not state a country.
source · checked 12 Oct 2026 · checked on the maker's own page
Status
Available
source · checked 12 Oct 2026 · checked on the maker's own page

What you can make with it

audio to text

Automatic punctuation, capitalisation, word and segment timestamps and language detection.
source · checked 12 Oct 2026 · checked on the maker's own page

Plans and prices

No paid plans listed.

source · checked 12 Oct 2026 · checked on the maker's own page

Rights and safety

Limits
max audio seconds: 1440; note: 24 minutes with full attention on an A100 80GB; up to 3 hours with local attention.
source · checked 12 Oct 2026 · checked on the maker's own page

Related

Where you can run it

Hugging Face

Other options for “A voice or speech, text from a recording”

Adobe Podcast (Enhance Speech), Adobe Premiere, AssemblyAI, Async (formerly Podcastle), AudioShake, Auphonic, Azure AI Speech, CAMB.AI, Cartesia, Cleanvoice AI, Clipchamp, Deepgram, Descript, Distil-Whisper, Dubverse, ElevenLabs, Filmora, Gling, Happy Scribe, Hugging Face, Kapwing, Krisp, MacWhisper, MiniMax Audio

About this entry

Facts last checked 12 Oct 2026 against the maker’s own pages (each fact links to its source). Entry status: draft. Plain-English summary written by Plainly.