Nemotron 3.5 Transcribe

A free NVIDIA model (October 2026) that turns recordings of up to an hour in 40 languages into text, with timestamps, who-said-what and translation. Commercial use is allowed under an open licence. Needs technical setup and a large NVIDIA GPU.

Go to Nemotron 3.5 Transcribe’s website

Free to try?Yes

Can I sell what I make?Yes

Open or closedDownloadable, with conditions

Skill neededTechnical

The facts, with sources

Open or closed: Downloadable, with conditions
You can download and run it, but the licence limits use
Weights downloadable (BF16, FP8, NVFP4).
source · checked 7 Nov 2026 · checked on the maker's own page
Licence: OpenMDW-1.1 (Permissive (e.g. Apache 2.0, MIT))
source · checked 7 Nov 2026 · checked on the maker's own page
Can I sell what I make?: Yes
Card: 'ready for commercial or non-commercial use'.
source · checked 7 Nov 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, Maker's own API
Runs with NeMo, vLLM, Transformers or SGLang; NVIDIA also lists a hosted NIM on build.nvidia.com (price not stated).
source · checked 7 Nov 2026 · checked, but the page was unclear or partial
Skill needed: Technical
source · checked 7 Nov 2026 · checked on the maker's own page
Made by: NVIDIA (US)
source · checked 7 Nov 2026 · checked on the maker's own page
Status: Available
Released 28 Oct 2026 on Hugging Face, NeMo, NGC and build.nvidia.com.
source · checked 7 Nov 2026 · checked on the maker's own page

What you can make with it

A voice or speech, text from a recording (main use)
A voice or speech, a written translation of a recording

audio to text

Also speaker-attributed transcripts, word timestamps, language ID and audio Q&A.

source · checked 7 Nov 2026 · checked on the maker's own page

Plans and prices

No paid plans listed.

source · checked 7 Nov 2026 · checked on the maker's own page

Rights and safety

Limits: max audio seconds: 3600; note: Recordings up to one hour.
source · checked 7 Nov 2026 · checked on the maker's own page

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, Granite Speech 5.0, Happy Scribe, Hugging Face, Kapwing, Krisp, LFM3-Audio

About this entry. Facts last checked 7 Nov 2026 against the maker’s own pages (each fact links to its source). Entry status: draft. Plain-English summary written by Plainly.

Spot an error? Tell us, with a link to the source. How corrections work