BroadNet Transcribe Arabic API
Updated
BroadNet Transcribe Arabic is a speech-to-text model based on Cohere Transcribe, served by BroadNet Inference at https://inference.broadnet.ai/v1/audio/transcriptions as broadnet/cohere-transcribe-arabic. It costs $0.0042 per audio minute, accepts WAV or MP3, and no audio or transcript is retained. It can use the starting credit.
At a glance
- Audio (speech to text)
- Arabic
- Arabic: Gulf, Egyptian, Levantine, Maghrebi and Modern Standard, per BroadNet
- WAV or MP3 up to 25 MiB per file
- OpenAI-compatible /v1/audio/transcriptions
- BroadNet reported #2 of 37 on the Open Universal Arabic ASR Leaderboard in September 2026; #4 of 39 when we checked on 15 November 2026
BroadNet's lab measurements: FLEURS-ar 3.8, MASC-ar 21.1 and dialect 23.4 word error rate (lower is better). The leaderboard rank of #2 of 37, and #1 among licences that permit API serving, is as reported by BroadNet in September 2026. When we checked the live leaderboard on 15 November 2026, it listed this model #4 of 39 by average word error rate (24.19), behind three newer entries.

| Model ID | broadnet/cohere-transcribe-arabic |
|---|---|
| Endpoint | /v1/audio/transcriptions |
| Supported inputs | Audio: WAV or MP3, up to 26,214,400 bytes (25 MiB) per file |
| Price | $0.0042 per audio minute |
| Languages | Arabic (language: ar) |
| Pay-as-you-go limits | 30 requests a minute, 4 in flight per key |
| Access | Starting credit eligible |
Prices on this page are from BroadNet's price feed snapshot dated , not checked live here. See official pricing for current rates.
official pricingReport a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What does BroadNet publish about BroadNet Transcribe Arabic?
From the BroadNet Transcribe Arabic page on broadnet.ai, checked 15 November 2026. Measurements are BroadNet's, on its hardware, unless a public leaderboard is named; we have not reproduced them.
| Model | Cohere Transcribe Arabic 07-2026, 2B parameters |
|---|---|
| Languages | Arabic: Gulf, Egyptian, Levantine, Maghrebi and Modern Standard |
| Input | WAV and MP3 audio |
| Output | Arabic script text; JSON or plain text |
| Hardware | NVIDIA GPUs operated by BroadNet in the United States |
| Licence | Apache-2.0 (Cohere Labs) |
BroadNet's measurements
- MASC real dialect speech, word error rate (lower is better): 21.1, against 22.3 for ElevenLabs Scribe v2 and 31.5 for OpenAI Whisper large-v3.
- Gulf, Egyptian and Levantine set: 23.4, against 25.6 and 31.8.
- FLEURS Arabic clean read speech: 3.8.
- 45 times realtime on one stream, 190 times batched, on Arabic audio.
- Open Universal Arabic ASR Leaderboard: #2 of 37 (average WER 25.87) and #1 among licences that permit API serving, as BroadNet reported in September 2026. When we checked the live board on 15 November 2026 it listed this model #4 of 39 (average WER 24.19), behind three newer entries.
Base weights: CohereLabs/cohere-transcribe-arabic-07-2026 on Hugging Face.
How BroadNet chooses and serves models: its GPU hardware and serving approach and the models it evaluated and ruled out.
Published benchmark results
BroadNet did not run these benchmarks; results are as published by the Cohere model card and may use different settings. BroadRouter is BroadNet's site. They refer to the published model CohereLabs/cohere-transcribe-arabic-07-2026; not measured on BroadNet's serving.
| Benchmark | Result | Setting or note | Source |
|---|---|---|---|
| Open Universal Arabic ASR Leaderboard: average WER | 25.87 | Word error rate, lower is better (CER 11.80); leaderboard as of 7 July 2026, as quoted on the card | Cohere model card, checked 2026-11-16 |
| Open Universal Arabic ASR Leaderboard: Common Voice WER | 5.82 | Word error rate, lower is better (CER 1.62); leaderboard as of 7 July 2026, as quoted on the card | Cohere model card, checked 2026-11-16 |
| Open Universal Arabic ASR Leaderboard: MASC clean WER | 19.60 | Word error rate, lower is better (CER 6.45); leaderboard as of 7 July 2026, as quoted on the card | Cohere model card, checked 2026-11-16 |
| Open Universal Arabic ASR Leaderboard: MASC noisy WER | 27.07 | Word error rate, lower is better (CER 10.13); leaderboard as of 7 July 2026, as quoted on the card | Cohere model card, checked 2026-11-16 |
| Open Universal Arabic ASR Leaderboard: MGB-2 WER | 15.54 | Word error rate, lower is better (CER 8.40); leaderboard as of 7 July 2026, as quoted on the card | Cohere model card, checked 2026-11-16 |
| Open Universal Arabic ASR Leaderboard: SADA WER | 37.47 | Word error rate, lower is better (CER 23.53); leaderboard as of 7 July 2026, as quoted on the card | Cohere model card, checked 2026-11-16 |
Side by side: BroadNet Transcribe Arabic vs BroadNet Transcribe, with prices, languages and monthly costs.
Call it in Python
Install the OpenAI SDK with pip install openai, export BROADNET_API_KEY, then run:
from openai import OpenAI
import os
client = OpenAI(base_url="https://inference.broadnet.ai/v1", api_key=os.environ["BROADNET_API_KEY"])
with open("voice-note.wav", "rb") as audio:
result = client.audio.transcriptions.create(
model="broadnet/cohere-transcribe-arabic",
file=audio,
language="ar", # always send the language
)
print(result.text)Or with curl
curl https://inference.broadnet.ai/v1/audio/transcriptions \
-H "Authorization: Bearer $BROADNET_API_KEY" \
-F model=broadnet/cohere-transcribe-arabic \
-F language=ar \
-F file=@voice-note.wavGuides for BroadNet Transcribe Arabic
Use BroadNet Transcribe Arabic in the console
The BroadNet console on inference.broadnet.ai manages accounts and keys. It does not receive your choice from this page, so:
- Sign in or create an account.
- Create a pay-as-you-go key; the starting credit covers this model.
- Send this model ID as
modelto/v1/audio/transcriptions:broadnet/cohere-transcribe-arabic
Open the BroadNet consoleEstimate the costRead the API reference
Compare BroadNet Transcribe Arabic
Questions
What is the model ID for BroadNet Transcribe Arabic?
Pass broadnet/cohere-transcribe-arabic as model to /v1/audio/transcriptions. The base URL is https://inference.broadnet.ai/v1.
How much does BroadNet Transcribe Arabic cost?
$0.0042 per audio minute, from your account balance.
Which languages does BroadNet Transcribe Arabic support?
Arabic. Always send the language field; without it Arabic can come back transliterated into Latin letters.
Can I use BroadNet Transcribe Arabic without a top-up?
Yes. It can use the starting credit for new accounts.
Where does BroadNet Transcribe Arabic run?
BroadNet's model page says: NVIDIA GPUs operated by BroadNet in the United States. Its privacy statement adds that processing locations for data-residency needs are agreed with BroadNet before you send data.
How does BroadNet Transcribe Arabic compare with Whisper and ElevenLabs?
On BroadNet's own benchmark it reports 21.1 word error rate on MASC dialect speech, against 22.3 for ElevenLabs Scribe v2 and 31.5 for Whisper large-v3 (lower is better). These are BroadNet's measurements; test on your own audio before you choose.
How does BroadNet Transcribe Arabic score on benchmarks?
As published, the Cohere model card reports 25.87 on Open Universal Arabic ASR Leaderboard: average WER, 5.82 on Open Universal Arabic ASR Leaderboard: Common Voice WER and 19.60 on Open Universal Arabic ASR Leaderboard: MASC clean WER (checked 2026-11-16). WER is word error rate: lower is better. BroadNet did not run these benchmarks, and sources may use different settings, so the figures are not directly comparable. Every published result we found, with its setting and source.
Is my data stored?
Inference content is not retained: audio and transcripts are processed for the request only. Necessary account, billing and security metadata, such as timings, token counts and cost, is retained as described in BroadNet's privacy statement. Read how.