Qwen3 ASR 1.7B
Qwen · released Jan 28, 2026 · Qwen/Qwen3-ASR-1.7B
Speech to textTranscribes a recording into words
- Type
- Open weightsApache License 2.0
- Languages
- 52
- Size
- 2.3B
Context measured in tokens
Our take
Written Sep 4, 2026Qwen3 ASR is a compact, Apache-licensed speech-to-text model that converts audio to text across 52 languages. At 2.04 billion parameters it is built for speed and broad language coverage rather than raw accuracy, though it still lands better than most models on several English audio conditions.
Pick this for fast transcription where throughput matters — it processes an hour of audio in roughly nine seconds on benchmark hardware. Use it for broad multilingual coverage across 52 languages, or for clean read-aloud and financial call transcription where it scores well. Skip it if you need commercial hosting, speaker separation, or the absolute lowest error rate on clean audio.
The case for it
- Extremely fast: 394 times real time, an hour of audio in roughly nine seconds on the benchmark hardware.
- Better than most models on every measured audio condition, from clean read-aloud to recorded meetings.
- Apache 2.0 licence with 52 languages supported, allowing unrestricted use and redistribution.
The case against it
- Not the most accurate option for clean audio despite its speed — roughly 38% more errors than the field leader on read-aloud.
- Accuracy degrades substantially on harder audio, from 1.24% on clean read-aloud to 9.88% on heavily accented international speech.
- No commercial hosting options currently available; you must self-host.
How good is it?
An open transcription model for turning podcasts and video audio into written text.
- transcribing podcasts and video audioPodcasts and video · 6th of 92
TranscriptionTurning speech into text
394×52nd of 74
an hour of audio in 9 seconds, on the board's own hardware. Your machine will differ.
52
Stated by the leaderboard; we do not hold the list itself.
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 model7 scoresEvery figure we hold, from 7 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. 20.1 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 28.1 GB spare means a 10% error in the size would not change the answer.
Apple M2 (8-core GPU, 8GB unified) · 8 GB
Room to spare. 3.3 GB spare means a 10% error in the size would not change the answer.
These cards answer whether Qwen3 ASR 1.7B 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 Qwen3 ASR 1.7B
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 allowsApache License 2.0, 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
Apache License 2.0
Fully permissive: commercial use, redistribution, and derivatives allowed. Requires attribution and a copy of the license. Includes an express patent grant.
Identifiers
- Hugging Face
- Qwen/Qwen3-ASR-1.7B
- Architecture
- Dense
- Takes in, gives back
- Audio in, text out
- Catalogue slug
- qwen-qwen3-asr-1-7b