--- license: other license_name: audarai-community-license-v1.0 license_link: https://www.audarai.com/license/audarai-community-license-v1.0/ language: - ar - en pipeline_tag: automatic-speech-recognition inference: false tags: - automatic-speech-recognition - asr - speech-recognition - arabic - arabic-asr - dialectal-arabic - emirati - gulf-arabic - streaming - realtime - gguf - llama-cpp - safetensors - transformers - audar ---
# Audar-ASR-V1-Turbo ยท GGUF ### Audar's Arabic-first speech-recognition model โ€” leaderboard-grade, dialect-aware. **From Arabic to the world.** ![License](https://img.shields.io/badge/license-AudarAI%20Community%20v1.0-6f42c1) ![Task](https://img.shields.io/badge/task-ASR-blue) ![Format](https://img.shields.io/badge/format-GGUF%20%2B%20safetensors-blue) ![Params](https://img.shields.io/badge/params-2.35B%20total-f59e0b) ![Open-AR-ASR](https://img.shields.io/badge/Open--AR--ASR-%231%20of%2037%20%C2%B7%2023.2%25%20avg%20WER-brightgreen) ![Avg CER](https://img.shields.io/badge/Avg%20CER-%231%20of%2037%20%C2%B7%209.2%25-brightgreen) ![Emirati](https://img.shields.io/badge/Emirati-19.4%25%20WER%20%2F%207.3%25%20CER-brightgreen) [![GitHub](https://img.shields.io/badge/GitHub-Audar--ASR--V1-181717?logo=github)](https://github.com/AudarAI/Audar-ASR-V1)

๐Ÿงญ Overview ยท ๐Ÿ“Š Benchmarks ยท ๐Ÿ’ป GGUF Deploy ยท ๐Ÿค— Transformers ยท ๐ŸŽ™๏ธ Streaming ยท โšก vLLM ยท ๐Ÿ“„ Tech Report ยท ๐Ÿ™ GitHub ยท โ˜๏ธ Audar API ยท ๐Ÿ“œ License

--- ## ๐Ÿงญ What it is **Audar-ASR-V1-Turbo** is an **Arabic-first generative speech-recognition model** โ€” the accuracy tier of the Audar-ASR family. It recasts transcription as **audio-conditioned next-token prediction** over a unified text vocabulary (a language-model decoder rather than a CTC or transducer objective), and is built on a permissively-licensed open-weight audio-LLM foundation and adapted in-house โ€” the contribution is the adaptation (the data curriculum and the alignment rubric), not the foundation: - ๐Ÿงฑ **Large-scale bilingual pretraining** โ€” 300,000+ hours of labeled audio, primarily Arabic and English, spanning MSA, Gulf, Egyptian, Levantine and Maghrebi speech, code-switching, and diverse acoustic channels. - ๐ŸŽฏ **Dialect-targeted fine-tuning** โ€” hardness sampling and multi-task conditioning focused on proper nouns, code-switching, and dialect-faithful orthography. - ๐Ÿง  **KTO preference alignment** โ€” Kahneman-Tversky Optimization on accented dialectal Arabic, with unpaired binary-desirability labels from trained native annotators across the Gulf, Levantine, Egyptian, and Maghrebi dialects, along five axes: verbatim accuracy, diacritic correctness, code-switch handling, named-entity preservation, and output formatting. The result is **state-of-the-art dialectal Arabic ASR** โ€” the lowest average WER *and* CER of any evaluated system on the *Open Universal Arabic ASR Leaderboard*. It transcribes MSA and every major Arabic dialect, code-switched Arabicโ€“English, and English, across **30 languages** in total. > Built on a **permissively-licensed open-weight audio-LLM foundation**; the adaptation, data, and > alignment are Audar's. Full method and results: [Audar-ASR-V1 Technical Report](https://github.com/AudarAI/Audar-ASR-V1/blob/main/report/Audar-ASR-V1-Technical-Report.pdf). ## Model summary
ModelAudar-ASR-V1-Turbo โ€” Arabic-first generative ASR (accuracy tier)
TaskAutomatic speech recognition (audio โ†’ text)
ApproachGenerative ASR โ€” audio encoder + language-model decoder (audio-conditioned next-token prediction)
Trainingbuilt on an open-weight audio-LLM foundation; adapted via a 4-stage curriculum โ€” 300k+ hrs bilingual pretraining โ†’ multi-task fine-tuning โ†’ dialect PEFT โ†’ KTO alignment
Decoder parameters2,031,739,904 (2.03B)
Audio encoder parameters317,477,504 (0.32B)
Total parameters2,349,217,408 (2.35B, bf16)
Audio input16 kHz mono; 30 s context (longer audio is chunked/streamed)
LanguagesArabic (MSA + Gulf/Egyptian/Levantine/Maghrebi dialects) + English + 28 more
RuntimeGGUF / llama.cpp โ€” CPU ยท GPU ยท edge
LicenseAudarAI Community License v1.0
## ๐Ÿ“Š Benchmarks Arabic dialectal ASR is **hard** โ€” heavily dialectal, conversational, code-switched speech is the frontier for every system. On the *Open Universal Arabic ASR Leaderboard*, **Audar-ASR-V1-Turbo ranks #1 of 37 systems** with the **lowest average WER (23.2 %) and the lowest average CER (9.2 %)** of any model evaluated โ€” and it is the single best system on **SADA**, **MASC-clean**, **MGB-2** and **Casablanca**. ### Open Universal Arabic ASR Leaderboard โ€” full standings *Per-dataset **WER %** across all six leaderboard test sets, plus the two composite averages. Lower is better; **Avg WER** is the ranking metric. Audar rows show the leaderboard maintainers' independent reproduction (Aug 2026) under the leaderboard's current normalization; other rows are as previously published by the leaderboard and may shift slightly when the full board is recomputed under the updated normalization. **Ours in bold.*** | # | Model | **Avg WER** | Avg CER | SADA | CV-18 | MASC-clean | MASC-noisy | MGB-2 | Casablanca | | --: | --- | --: | --: | --: | --: | --: | --: | --: | --: | | **1** | **Audar-ASR-V1-Turbo (Ours)** | **23.17** | **9.20** | **28.92** | 8.09 | **16.73** | 27.19 | **11.08** | **47.02** | | 2 | CohereLabs/cohere-transcribe-arabic-07-2026 | 25.87 | 11.80 | 37.47 | **5.82** | 19.60 | **27.07** | 15.54 | 49.71 | | 3 | omnilingual-asr/omniASR_LLM_7B | 28.32 | 12.52 | 41.61 | 8.75 | 19.69 | 29.29 | 14.13 | 56.46 | | 4 | omnilingual-asr/omniASR_LLM_3B | 29.96 | 13.77 | 46.18 | 9.15 | 19.90 | 30.03 | 14.22 | 60.27 | | 5 | omnilingual-asr/omniASR_LLM_1B | 29.96 | 13.40 | 43.84 | 9.55 | 20.03 | 30.26 | 15.34 | 60.68 | | 6 | CohereLabs/cohere-transcribe-03-2026 | 30.67 | 16.37 | 60.11 | 8.17 | **8.66** | **19.01** | 25.33 | 62.71 | | 7 | Qwen/Qwen3-Omni-30B-A3B-Instruct | 30.71 | 13.67 | 44.82 | 11.46 | 21.47 | 30.85 | 13.09 | 62.55 | | 8 | nvidia-conformer-ctc-large-arabic (lm) | 32.91 | 13.84 | 44.52 | 8.80 | 23.74 | 34.29 | 17.20 | 68.90 | | 9 | omnilingual-asr/omniASR_LLM_300M | 32.96 | 14.84 | 51.38 | 12.03 | 20.66 | 32.45 | 16.58 | 64.64 | | 10 | google/gemma-4-E4B-it | 32.98 | 13.71 | 43.40 | 19.65 | 24.86 | 33.59 | 17.72 | 58.63 | | 11 | Qwen/Qwen3-ASR-1.7B | 33.36 | 12.33 | 45.53 | 16.90 | 24.37 | 34.29 | 16.57 | 64.47 | | 12 | mistralai/Voxtral-Small-24B-2507 | 34.47 | 15.29 | 50.82 | 15.25 | 23.96 | 34.43 | 16.03 | 66.30 | | 13 | nvidia-conformer-ctc-large-arabic (greedy) | 34.74 | 13.37 | 47.26 | 10.60 | 24.12 | 35.64 | 19.69 | 71.13 | | 14 | google/gemma-4-E2B-it | 35.87 | 15.34 | 46.23 | 23.76 | 27.47 | 36.15 | 20.72 | 60.87 | | 15 | openai/whisper-large-v3 | 36.86 | 17.21 | 55.96 | 17.83 | 24.66 | 34.63 | 16.26 | 71.81 | | 16 | omnilingual-asr/omniASR_CTC_3B | 37.78 | 19.79 | 69.85 | 14.19 | 21.48 | 34.60 | 18.96 | 67.58 | | 17 | omnilingual-asr/omniASR_CTC_7B | 38.12 | 20.91 | 72.69 | 12.47 | 21.08 | 35.04 | 20.43 | 67.02 | | 18 | facebook/seamless-m4t-v2-large | 38.16 | 17.03 | 62.52 | 21.70 | 25.04 | 33.24 | 20.23 | 66.25 | | 19 | omnilingual-asr/omniASR_CTC_1B | 39.29 | 20.47 | 71.42 | 17.55 | 22.76 | 35.73 | 19.96 | 68.32 | | 20 | openai/whisper-large-v3-turbo | 40.05 | 18.87 | 60.36 | 25.73 | 25.51 | 37.16 | 17.75 | 73.79 | | 21 | openai/whisper-large-v2 | 40.20 | 19.55 | 57.46 | 21.77 | 27.25 | 38.55 | 25.17 | 71.01 | | 22 | Qwen/Qwen3-ASR-0.6B | 42.19 | 16.23 | 53.75 | 28.28 | 31.34 | 42.63 | 25.45 | 71.68 | | 23 | openai/whisper-large | 42.57 | 20.49 | 63.24 | 26.04 | 28.89 | 40.79 | 24.28 | 72.18 | | 24 | mistralai/Voxtral-Mini-3B-2507 | 42.58 | 19.90 | 63.65 | 22.12 | 28.37 | 41.27 | 22.56 | 77.52 | | 25 | asafaya/hubert-large-arabic-transcribe | 45.50 | 17.35 | 67.82 | 8.01 | 32.94 | 50.16 | 37.51 | 76.53 | | 26 | openai/whisper-medium | 45.57 | 22.27 | 67.71 | 28.07 | 29.99 | 42.91 | 29.32 | 75.44 | | 27 | nvidia-Parakeet-ctc-1.1b-concat | 46.54 | 23.88 | 70.70 | 26.34 | 30.49 | 45.95 | 24.94 | 80.80 | | 28 | omnilingual-asr/omniASR_CTC_300M | 46.65 | 21.86 | 78.11 | 27.90 | 28.40 | 43.26 | 26.85 | 75.35 | | 29 | nvidia-Parakeet-ctc-1.1b-universal | 51.96 | 25.19 | 73.58 | 40.01 | 36.16 | 50.03 | 30.68 | 81.30 | | 30 | microsoft/VibeVoice-ASR | 52.99 | 28.95 | 69.83 | 44.25 | 32.95 | 52.43 | 25.10 | 93.37 | | 31 | facebook/mms-1b-all | 54.54 | 21.45 | 77.48 | 26.52 | 38.82 | 57.33 | 39.16 | 87.95 | | 32 | openai/whisper-small | 55.13 | 21.68 | 78.02 | 24.18 | 35.93 | 56.36 | 48.64 | 87.64 | | 33 | whitefox123/w2v-bert-2.0-arabic-4 | 58.13 | 27.62 | 87.34 | 41.79 | 37.82 | 53.28 | 40.66 | 87.88 | | 34 | jonatasgrosman/wav2vec2-large-xlsr-53-arabic | 60.98 | 25.61 | 86.82 | 23.00 | 42.75 | 64.27 | 56.29 | 92.72 | | 35 | speechbrain/asr-wav2vec2-commonvoice-14-ar | 65.74 | 30.93 | 88.54 | 29.17 | 49.10 | 69.57 | 64.37 | 93.68 | **Bold** = best in column. The 37th system, our sibling edge model [**Audar-ASR-V1-Flash** (0.78B)](https://huggingface.co/audarai/Audar-ASR-V1-Flash), enters at 32.04 avg WER โ€” see its card for the full row. Audar-ASR-V1-Turbo owns both composite averages and leads on SADA, MASC-clean, MGB-2 and Casablanca; the recent Cohere and OmniASR systems are the closest competitors, each strongest on a subset of the conversational and clean-read sets. Casablanca (Moroccan Darija) is the hardest set for every system. ### Emirati Arabic | Set | WER % | CER % | |---|---|---| | **Emirati** (Mixat, full 1,585-clip test) | **19.4** | **7.3** | On Emirati, the **real recognition error is โ‰ˆ 7.3 %** โ€” near-parity with spontaneous English โ€” while the residual up to 19.4 % WER is largely **orthographic convention** (near-miss spelling of the *same* word, e.g. ุงู†ุชูˆโ†”ุงู†ุชูˆุง, and Latin-vs-Arabic rendering of English loanwords), not misrecognition. ## ๐Ÿ Benchmark-parity inference (qwen-asr) โ€” recommended Our leaderboard numbers were produced with the [`qwen-asr`](https://pypi.org/project/qwen-asr/) package, which implements this model's I/O protocol natively โ€” and were independently reproduced by the leaderboard maintainers with this exact code: ```python # pip install qwen-asr torch import torch from qwen_asr import Qwen3ASRModel model = Qwen3ASRModel.from_pretrained( "audarai/Audar-ASR-V1-Turbo", dtype=torch.bfloat16, device_map="cuda:0", max_inference_batch_size=16, max_new_tokens=256, ) results = model.transcribe(audio=["clip.wav"], language=["Arabic"]) print(results[0].text) ``` Protocol handling is **mandatory**, not optional: - `language="Arabic"` makes the package prefill `language Arabic` into the prompt, so the model never free-runs language identification. - The model's no-speech verdict (`language None`) is mapped to an empty transcript; without this, non-speech audio (music, silence) can produce repetition loops. - `max_new_tokens=256` and bf16 are the exact decode settings behind our published numbers. If you use raw `transformers` (below), you **must** strip the `language ` output prefix yourself and expect degraded scores on non-speech-heavy data. ## ๐Ÿ’ป GGUF inference (llama.cpp) Turbo runs on **llama.cpp** via the multimodal (`mtmd`) path โ€” a quantized **decoder** GGUF plus a **BF16 audio projector** (`mmproj`). Build a recent llama.cpp (with Qwen3-ASR support), then: ```bash ./llama-mtmd-cli \ -m Audar-ASR-V1-Turbo-Q8_0.gguf \ --mmproj mmproj-Audar-ASR-V1-Turbo.gguf \ --audio clip.wav \ -sys "ูุฑู‘ุบ ุงู„ูƒู„ุงู… ุงู„ุนุฑุจูŠ ุงู„ุชุงู„ูŠ." \ --temp 0 ``` > โš ๏ธ The **audio projector (`mmproj`) must stay BF16** (its `ClippableLinear` is numerically > sensitive). The **decoder** quantizes normally. Prefer a managed endpoint? The Audar-ASR family is also available via the [**Audar API/SDK**](https://www.audarai.com) โ€” streaming, speaker-attributed transcription, and diarization, production-hosted. ### GGUF variants | File | Approx. size | Notes | |---|---|---| | `Audar-ASR-V1-Turbo-Q4_K_M.gguf` | ~1.28 GB | Smallest; constrained hardware | | `Audar-ASR-V1-Turbo-Q8_0.gguf` | ~2.16 GB | Near-lossless (recommended) | | `Audar-ASR-V1-Turbo.gguf` (BF16) | ~4.07 GB | Full precision decoder | | `mmproj-Audar-ASR-V1-Turbo.gguf` | ~0.64 GB | **BF16 audio encoder โ€” required, keep BF16** | ## ๐Ÿค— Transformers (full-precision safetensors) The **full-precision bf16 weights** are published at the **repo root** โ€” the reference checkpoint the GGUF and W4A16 builds are derived from (**2,349,217,408** params, `safetensors`). Standard ๐Ÿค— Transformers, loaded with `trust_remote_code=True` (the repo ships the self-contained Qwen3-ASR code). ```python # pip install "transformers==4.57.6" torch librosa import torch, librosa from transformers import AutoProcessor, AutoModelForCausalLM repo = "audarai/Audar-ASR-V1-Turbo" proc = AutoProcessor.from_pretrained(repo, trust_remote_code=True) model = AutoModelForCausalLM.from_pretrained( repo, trust_remote_code=True, dtype=torch.bfloat16, device_map="cuda:0", ).eval() SYSTEM = "ูุฑู‘ุบ ุงู„ูƒู„ุงู… ุงู„ุนุฑุจูŠ ุงู„ุชุงู„ูŠ." # "Transcribe the following Arabic speech." audio, _ = librosa.load("clip.wav", sr=16000, mono=True) conv = [{"role": "system", "content": SYSTEM}, {"role": "user", "content": [{"type": "audio"}]}] # audio placeholder (a list, not "