ibm-granite/granite-4.0-1b-speech

`ibm-granite/granite-4.0-1b-speech` powered by vLLM - [Original Model Card](https://huggingface.co/ibm-granite/granite-4.0-1b-speech) - [vLLM Documentation](https://docs.vllm.ai/) ## Transcriptions A…

azureFoundry/HuggingFace/ibm-granite-granite-4.0-1b-speech@1

ibm-granite/granite-4.0-1b-speech powered by vLLM

Transcriptions API

Send Request

You can use cURL or any REST Client to send a request to the Azure ML endpoint with your Azure ML token.

curl <AZUREML_ENDPOINT_URL> \
    -X POST \
    -H "Authorization: Bearer <AZUREML_TOKEN>" \
    -H "Content-Type: multipart/form-data" \
    -F "model="ibm-granite/granite-4.0-1b-speech" \
    -F file="@/path/to/file/audio.mp3"

Supported Parameters

The following are the only mandatory parameters to send in the HTTP POST request to v1/audio/transcriptions.

  • model (string): Model ID used to generate the response, in this case since only a single model is deployed within the same endpoint you can either set it to ibm-granite/granite-4.0-1b-speech or leave it blank instead.
  • file (file): The audio file object (not file name) to transcribe, in one of these formats: flac, mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.

The rest of the parameters are optional, and since this model is powered by vLLM with an OpenAI compatible interface on top for the Audio Transcriptions API for automatic speech recognition, then the I/O interfaces for both transcription and streaming are the same as in OpenAI Audio Transcriptions API.

You can find the whole specification of the allowed parameters in the OpenAI Audio Transcriptions API Specification, or alternatively in the endpoint /openapi.json for the current Azure ML Endpoint.

Specifications & capabilities

Input
audio
Output
text
Published model ID
granite-4.0-1b-speech
Capabilities
audioInput

Evaluations

21–37 / 37

Open ASR Leaderboard

514.678

source id
openASR
leaderboard name
Voice Arena Monsoon RTFx
dimension
runtime
metric
Voice Arena Monsoon RTFx
rank
28
aggregation scope
status
published
direction
higherIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
Voice Arena Monsoon RTFx
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
49
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

666.99

source id
openASR
leaderboard name
RTFx
dimension
runtime
metric
RTFx
rank
29
aggregation scope
status
published
direction
higherIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
RTFx
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
49
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

584.542

source id
openASR
leaderboard name
LS Other RTFx
dimension
runtime
metric
LS Other RTFx
rank
29
aggregation scope
status
published
direction
higherIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
LS Other RTFx
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
49
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

671.31

source id
openASR
leaderboard name
English · RTFx
dimension
runtime
metric
English · RTFx
rank
32
aggregation scope
status
published
direction
higherIsBetter
conditions
view
english
suite
OpenASR official public table
datasets
  • AMI-Cleaned
  • Earnings22-Cleaned-AA-chunked
  • Gigaspeech-Cleaned
  • LS Clean
  • LS Other
  • SPGISpeech
  • URGENT24
  • Voxpopuli-AA-Cleaned
  • URGENT24 Clean
metric_method
publisher WER/RTFx for selected public view
space_revision
3b5317f4e49b6cb40502b95065b29fbfe08284ce
publisher_version
13-11-2026
metric_kind
English · RTFx
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
public/english.json
published_at
2026-11-13T11:42:34Z
source_revision
8f2cc9c9f3b01b8670c8640efc67f24a3ea076c2bf5e90e7c674a1a829d69bd7
metric_population
49
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech 🤗
Original result ↗

Open ASR Leaderboard

317.294

source id
openASR
leaderboard name
Voxpopuli-AA-Cleaned RTFx
dimension
runtime
metric
Voxpopuli-AA-Cleaned RTFx
rank
36
aggregation scope
status
published
direction
higherIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
Voxpopuli-AA-Cleaned RTFx
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
49
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

238.798

source id
openASR
leaderboard name
Earnings22-Cleaned-AA-chunked RTFx
dimension
runtime
metric
Earnings22-Cleaned-AA-chunked RTFx
rank
36
aggregation scope
status
published
direction
higherIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
Earnings22-Cleaned-AA-chunked RTFx
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
49
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

5.9%

source id
openASR
leaderboard name
English · Average WER
dimension
speech
metric
English · Average WER
rank
37
aggregation scope
status
published
direction
lowerIsBetter
conditions
view
english
suite
OpenASR official public table
datasets
  • AMI-Cleaned
  • Earnings22-Cleaned-AA-chunked
  • Gigaspeech-Cleaned
  • LS Clean
  • LS Other
  • SPGISpeech
  • URGENT24
  • Voxpopuli-AA-Cleaned
  • URGENT24 Clean
metric_method
publisher WER/RTFx for selected public view
space_revision
3b5317f4e49b6cb40502b95065b29fbfe08284ce
publisher_version
13-11-2026
metric_kind
English · Average WER
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
public/english.json
published_at
2026-11-13T11:42:34Z
source_revision
8f2cc9c9f3b01b8670c8640efc67f24a3ea076c2bf5e90e7c674a1a829d69bd7
metric_population
64
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech 🤗
Original result ↗

Open ASR Leaderboard

7.02%

source id
openASR
leaderboard name
avg (WER)
dimension
speech
metric
avg (WER)
rank
42
aggregation scope
status
published
direction
lowerIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
avg (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
65
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

8.44%

source id
openASR
leaderboard name
Gigaspeech-Cleaned WER (WER)
dimension
speech
metric
Gigaspeech-Cleaned WER (WER)
rank
43
aggregation scope
status
published
direction
lowerIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
Gigaspeech-Cleaned WER (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
65
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

8.44%

source id
openASR
leaderboard name
English · Gigaspeech-Cleaned (WER)
dimension
speech
metric
English · Gigaspeech-Cleaned (WER)
rank
43
aggregation scope
status
published
direction
lowerIsBetter
conditions
view
english
suite
OpenASR official public table
datasets
  • AMI-Cleaned
  • Earnings22-Cleaned-AA-chunked
  • Gigaspeech-Cleaned
  • LS Clean
  • LS Other
  • SPGISpeech
  • URGENT24
  • Voxpopuli-AA-Cleaned
  • URGENT24 Clean
metric_method
publisher WER/RTFx for selected public view
space_revision
3b5317f4e49b6cb40502b95065b29fbfe08284ce
publisher_version
13-11-2026
metric_kind
English · Gigaspeech-Cleaned (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
public/english.json
published_at
2026-11-13T11:42:34Z
source_revision
8f2cc9c9f3b01b8670c8640efc67f24a3ea076c2bf5e90e7c674a1a829d69bd7
metric_population
64
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech 🤗
Original result ↗

Open ASR Leaderboard

4.84%

source id
openASR
leaderboard name
Voice Arena Monsoon WER (WER)
dimension
speech
metric
Voice Arena Monsoon WER (WER)
rank
45
aggregation scope
status
published
direction
lowerIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
Voice Arena Monsoon WER (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
65
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

8.53%

source id
openASR
leaderboard name
Earnings22-Cleaned-AA-chunked WER (WER)
dimension
speech
metric
Earnings22-Cleaned-AA-chunked WER (WER)
rank
46
aggregation scope
status
published
direction
lowerIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
Earnings22-Cleaned-AA-chunked WER (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
65
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

8.53%

source id
openASR
leaderboard name
English · Earnings22-Cleaned-AA-chunked (WER)
dimension
speech
metric
English · Earnings22-Cleaned-AA-chunked (WER)
rank
47
aggregation scope
status
published
direction
lowerIsBetter
conditions
view
english
suite
OpenASR official public table
datasets
  • AMI-Cleaned
  • Earnings22-Cleaned-AA-chunked
  • Gigaspeech-Cleaned
  • LS Clean
  • LS Other
  • SPGISpeech
  • URGENT24
  • Voxpopuli-AA-Cleaned
  • URGENT24 Clean
metric_method
publisher WER/RTFx for selected public view
space_revision
3b5317f4e49b6cb40502b95065b29fbfe08284ce
publisher_version
13-11-2026
metric_kind
English · Earnings22-Cleaned-AA-chunked (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
public/english.json
published_at
2026-11-13T11:42:34Z
source_revision
8f2cc9c9f3b01b8670c8640efc67f24a3ea076c2bf5e90e7c674a1a829d69bd7
metric_population
64
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech 🤗
Original result ↗

Open ASR Leaderboard

4.41%

source id
openASR
leaderboard name
English · Voxpopuli-AA-Cleaned (WER)
dimension
speech
metric
English · Voxpopuli-AA-Cleaned (WER)
rank
52
aggregation scope
status
published
direction
lowerIsBetter
conditions
view
english
suite
OpenASR official public table
datasets
  • AMI-Cleaned
  • Earnings22-Cleaned-AA-chunked
  • Gigaspeech-Cleaned
  • LS Clean
  • LS Other
  • SPGISpeech
  • URGENT24
  • Voxpopuli-AA-Cleaned
  • URGENT24 Clean
metric_method
publisher WER/RTFx for selected public view
space_revision
3b5317f4e49b6cb40502b95065b29fbfe08284ce
publisher_version
13-11-2026
metric_kind
English · Voxpopuli-AA-Cleaned (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
public/english.json
published_at
2026-11-13T11:42:34Z
source_revision
8f2cc9c9f3b01b8670c8640efc67f24a3ea076c2bf5e90e7c674a1a829d69bd7
metric_population
64
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech 🤗
Original result ↗

Open ASR Leaderboard

4.41%

source id
openASR
leaderboard name
Voxpopuli-AA-Cleaned WER (WER)
dimension
speech
metric
Voxpopuli-AA-Cleaned WER (WER)
rank
52
aggregation scope
status
published
direction
lowerIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
Voxpopuli-AA-Cleaned WER (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
65
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Open ASR Leaderboard

3.55%

source id
openASR
leaderboard name
English · SPGISpeech (WER)
dimension
speech
metric
English · SPGISpeech (WER)
rank
53
aggregation scope
status
published
direction
lowerIsBetter
conditions
view
english
suite
OpenASR official public table
datasets
  • AMI-Cleaned
  • Earnings22-Cleaned-AA-chunked
  • Gigaspeech-Cleaned
  • LS Clean
  • LS Other
  • SPGISpeech
  • URGENT24
  • Voxpopuli-AA-Cleaned
  • URGENT24 Clean
metric_method
publisher WER/RTFx for selected public view
space_revision
3b5317f4e49b6cb40502b95065b29fbfe08284ce
publisher_version
13-11-2026
metric_kind
English · SPGISpeech (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
public/english.json
published_at
2026-11-13T11:42:34Z
source_revision
8f2cc9c9f3b01b8670c8640efc67f24a3ea076c2bf5e90e7c674a1a829d69bd7
metric_population
64
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech 🤗
Original result ↗

Open ASR Leaderboard

3.55%

source id
openASR
leaderboard name
SPGISpeech WER (WER)
dimension
speech
metric
SPGISpeech WER (WER)
rank
55
aggregation scope
status
published
direction
lowerIsBetter
conditions
suite
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
metric_method
published WER/RTFx
metric_kind
SPGISpeech WER (WER)
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
source_file
hf-audio/open-asr-leaderboard-results/english_short_latest.csv
published_at
2026-10-21T14:20:31Z
source_revision
943867e1b478153a2161b0522c130917d0d96b01a780660e00252b573bc2cf13
metric_population
65
benchmark_evaluation
true
source model id
ibm-granite/granite-4.0-1b-speech
Original result ↗

Provider specifications & access

1–1 / 1

ibm-granite/granite-4.0-1b-speech ↗

source model id
HuggingFace/ibm-granite-granite-4.0-1b-speech@1
hosting
foundry-managed-compute
platforms
  • Microsoft Foundry
license
apache-2.0
architecture
input_modalities
  • audio
output_modalities
  • text
capabilities

    Safety assessments

    1–1 / 1

    Foundation Model Transparency Index 2025 ↗

    95 / 100

    source id
    stanfordFMTI
    canonical model id
    azureFoundry/HuggingFace/ibm-granite-granite-4.0-1b-speech@1
    taxonomy
    transparency
    hazard category
    transparency
    hazard name
    Organization transparency
    policy area
    IBM
    assessment type
    organization-level disclosure
    metric
    disclosure_score
    result label
    95 / 100
    sample size
    100
    updated at
    2026-11-09T18:08:05.227Z

    Original sources

    1–6 / 6

    Hugging Face · ibm-granite/granite-4.0-1b-speech ↗

    Microsoft Foundry · Model page ↗

    Microsoft Foundry · Model documentation ↗

    GitHub · Catalog ↗

    Hugging Face · Catalog ↗

    Microsoft Foundry · Catalog ↗

    Source links

    1–3 / 3

    Explore more models

    Models

    Loading
    EVIDENCE / COMPARISON

    Benchmarks

    Explore every benchmark

    Loading

    These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

    Install TensorTrace