Niagara 9m Batch.en
Applied Brain Research · released Sep 12, 2026 · abr-ai/niagara-9m-batch.en
Speech to textTranscribes a recording into words
- Type
- Open weightsCustom licence
- Languages
- 1
- Size
- 9M
Context measured in tokens · download allowed, licence restricts use
Our take
Written Sep 28, 2026Niagara 9m Batch.en is a speech-to-text model you can download and run yourself, and it is built for one thing: turning a large audio backlog into rough text as fast as possible. Its accuracy is among the weakest we list, so it suits bulk work you will clean up rather than anything you would publish unchecked.
Use it for bulk transcription where a rough draft is enough — search indexing, keyword spotting, or a first pass you will tidy afterwards — and for running on modest hardware, since it is a small download. The licence puts conditions on commercial use and redistribution, so it needs reading before you build on it. Skip it if you need accurate transcription of meetings or accented speech, or if you want a host to run it for you.
The case for it
- 7,158 times real time on the leaderboard's own hardware, an hour of audio in under a second, so a large backlog becomes practical to work through.
- You can download it and run it yourself, and at 9 million parameters in total it is a small download for modest hardware.
The case against it
- 13.5% of words wrong on average across nine English test sets — roughly one word in seven — against a field middle of 5.2% across 76 models.
- Worse than most models on every condition measured: 21.5% of words wrong on accented speech and 20.9% in recorded meetings, against a field middle of 7.9% and 10.3%.
- English only, and no source we hold measures any other language, so accuracy elsewhere is unverified in our data.
How good is it?
An open speech-to-text model for turning recordings into written text, though it trails most others on accuracy.
- turning spoken English into written textOpen ASR WER · 75th of 76
- transcribing recordings of meetings in a roomRecorded meetings · 87th of 92
- transcribing speakers with a range of accentsAccented speech · 75th of 76
- transcribing podcasts and video audioPodcasts and video · 89th of 92
- transcribing clear recordings of people reading aloudClean read speech · 91st of 92
TranscriptionTurning speech into text1 of 5Open ASR WER · 75th of 76
86.5%
Misses roughly one word in 7, averaged over nine English test sets.
7,158×7th of 74
an hour of audio in under a second, on the board's own hardware. Your machine will differ.
1
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; the placing beneath each rate is against every model measured on that set.
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.
Each of these is the same transcription job on a different kind of recording, so together they say where it holds up and where it slips — not how closely it follows an instruction.
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?
Comfortable fit
GeForce RTX 4090 · 24 GB
Room to spare. 21.6 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. 22.8 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. 4.8 GB spare means a 10% error in the size would not change the answer.
These cards answer whether Niagara 9m Batch.en loads, not how fast it transcribes. Throughput figures for a transcription model come from its text decoder, so treat this as a fit answer rather than a speed one.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first. This is a fit answer: whether it loads, not how fast it transcribes.
Models people weigh against Niagara 9m Batch.en
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
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 allowsCustom licence, 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
Custom licence
This model ships custom license terms that don't map to a known template. We haven't parsed them, so commercial use, redistribution and derivatives are unverified — review the original terms before shipping.
Identifiers
- Hugging Face
- abr-ai/niagara-9m-batch.en
- Architecture
- Dense
- Takes in, gives back
- Audio in, text out
- Catalogue slug
- abr-ai-niagara-9m-batch-en