Nemotron 3.5 ASR Streaming 0.6b
NVIDIA · nvidia/nemotron-3.5-asr-streaming-0.6b
Speech to textTranscribes a recording into words
- Type
- Open weightsopenmdw-1.1
- Languages
- 35
- Size
- 0.6B
Context measured in tokens
Our take
Written Aug 2, 2026Nemotron 3.5 ASR Streaming is a tiny downloadable speech-to-text model from NVIDIA that turns audio into written words across 35 languages. It is extremely fast on benchmark hardware, processing an hour of audio in roughly two seconds, though its accuracy varies sharply between clean and difficult recordings.
Pick this for batch transcription jobs where speed matters more than perfect accuracy, or for multilingual deployments needing 35 languages from a single small model. Use it if you want to self-host without relying on an API. Skip it if you need hosted availability, if your audio is messy meetings or heavily accented speech, or if you require a verified release date.
The case for it
- Processes one hour of audio in approximately 2.4 seconds on benchmark hardware, at 1,489.58 times real time.
- Strong on clean read-aloud audio, with 2.83% of words wrong.
- Resilient on financial calls and European-accented speech, at 3.27% and 4.24% respectively — both well below its own average.
- 35 languages from a 0.6B-parameter model, including English, Spanish, German, French, Italian, Arabic, Japanese and Korean.
The case against it
- Accuracy collapses on challenging real-world audio: recorded meetings hit 13.42% and accented speech reaches 14.95%, more than five times its clean-read rate.
- No hosted availability in our catalogue; you must self-host.
- Release date is undisclosed in our data.
How good is it?
TranscriptionTurning speech into text1 of 5Open ASR WER · 66th of 74
92.1%
Misses roughly one word in 13, averaged over nine English test sets.
1,490×22nd of 62
an hour of audio in 2 seconds, on the board's own hardware. Your machine will differ.
35
Listed on the model card. The accuracy above is English only.
Percentage of words wrong on each set, lower better. Bars are scaled to this model's own worst case, not to the board.
Which languages ↓ ↑
English · Spanish · German · French · Italian · Arabic · Japanese · Korean · Portuguese · Russian · Hindi · Chinese · Vietnamese · Hebrew · Dutch · Czech · Danish · Polish · Norwegian · Swedish · Thai · Turkish · Bulgarian · Greek · Estonian · Finnish · Croatian · Hungarian · Lithuanian · Latvian · Romanian · Slovak · Ukrainian · Maltese · Slovenian
The figures above come from the Open ASR Leaderboard, an independent public test that runs every model on the same recordings. It is the only measurement of transcription quality we know of, so there are no other scores to show.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done, which is why they get no rating.
Every published score for this model9 scoresEvery figure we hold, from 9 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
- Fits in memory
- weights load entirely on the card
- Spills to system RAM
- some weights offload; much slower
- Too large
- will not load even with offload
- est
- size is calculated; the verdict could change by 10%
Comfortable fit
GeForce RTX 4090 · 24 GB
Room to spare. 21.2 GB spare means a 10% error in the size would not change the answer.
Apple M1 Pro (16-core GPU) · 32 GB
Room to spare. 22.4 GB spare means a 10% error in the size would not change the answer.
Apple M1 (8-core GPU, 8GB unified) · 8 GB
Room to spare. 4.4 GB spare means a 10% error in the size would not change the answer.
Memory use by level
Against a 24 GB card.
All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.
Check against your own machine → · All 71 devices, with every size →
When we formed this view
Dates behind this page
Prices last checked 6h ago
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
Licence and identifiers
What the licence allowsopenmdw-1.1, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
openmdw-1.1
License tag "openmdw-1.1" imported from Hugging Face; terms pending curation.
Identifiers
- Hugging Face
- nvidia/nemotron-3.5-asr-streaming-0.6b
- Modality record
- audio->text
- Catalogue slug
- nvidia-nemotron-3-5-asr-streaming-0-6b