Fact · Speech recognition
56data points
The word error rates of speech recognition models and APIs on the Open ASR Leaderboard's public English test sets (tracked weekly) dataset holds 56 data points, each a published figure traced to its source.
Calculated from 56 data points on October 20, 2026
- How it is calculated
- Count of the published facts that are rows of the dataset: 56.
A statistic of the dataset Open ASR Leaderboard: Speech Recognition Accuracy, Tracked, calculated by code from its published rows.
Calculated from
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), abr-ai/niagara-19m-batch.en had an average word error rate of 12.08 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), abr-ai/niagara-38m-batch.en had an average word error rate of 10.31 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), abr-ai/niagara-84m-batch.en had an average word error rate of 8.87 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), abr-ai/niagara-9m-batch.en had an average word error rate of 15.27 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), assemblyai/universal-3-5-pro had an average word error rate of 4.87 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), bosonai/higgs-audio-v3-stt had an average word error rate of 5.22 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), CohereLabs/cohere-transcribe-03-2026 had an average word error rate of 5.21 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), distil-whisper/distil-large-v3.5 had an average word error rate of 6.18 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), Edge0/ARK-ASR-0.6B had an average word error rate of 5.44 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), Edge0/ARK-ASR-3B had an average word error rate of 5.45 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), Edge0/Audio8-ASR-0.1B had an average word error rate of 7.84 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), elevenlabs/scribe_v2 had an average word error rate of 4.55 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), espnet/owsm_ctc_v4_1B had an average word error rate of 7.37 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), facebook/omniASR-CTC-7B-v2 had an average word error rate of 10.13 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), facebook/omniASR-LLM-7B-v2 had an average word error rate of 7.34 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), FermionResearch/Phonon-2 had an average word error rate of 6.25 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), FunAudioLLM/Fun-ASR-Nano-2512-hf had an average word error rate of 6.65 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), gladia/solaria-3 had an average word error rate of 5.59 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), google/gemma-4-E4B-it had an average word error rate of 11.38 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), HojoAI/Hojo-ASR-V1 had an average word error rate of 5.22 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), ibm-granite/granite-4.0-1b-speech had an average word error rate of 5.86 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), ibm-granite/granite-speech-4.1-2b had an average word error rate of 5.35 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), ibm-granite/granite-speech-4.1-2b-nar had an average word error rate of 5.40 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), ibm-granite/granite-speech-5.0-470m-turboctc had an average word error rate of 5.78 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), ibm-granite/granite-speech-5.0-470m-turboctc-nc had an average word error rate of 5.44 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), meta/muse-voice-transcribe had an average word error rate of 5.75 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), microsoft/azure-speech-07-2026 had an average word error rate of 4.49 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), microsoft/VibeVoice-ASR-HF had an average word error rate of 7.12 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), mistralai/Voxtral-Mini-3B-2507 had an average word error rate of 6.66 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), mistralai/Voxtral-Mini-4B-Realtime-2602 had an average word error rate of 7.74 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), mistralai/Voxtral-Small-24B-2507 had an average word error rate of 5.81 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), modulate/multilingual had an average word error rate of 4.53 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/canary-1b-v2 had an average word error rate of 6.69 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/canary-qwen-2.5b had an average word error rate of 5.38 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/nemotron-3.5-asr-streaming-0.6b had an average word error rate of 9.24 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/nemotron-speech-streaming-en-0.6b had an average word error rate of 6.06 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/parakeet-ctc-1.1b had an average word error rate of 6.92 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/parakeet-tdt-0.6b-v2 had an average word error rate of 5.45 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/parakeet-tdt-0.6b-v2 (fast-gpu-asr) had an average word error rate of 5.44 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/parakeet-tdt-0.6b-v3 had an average word error rate of 5.58 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), nvidia/parakeet-tdt-0.6b-v3 (fast-gpu-asr) had an average word error rate of 5.46 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), openai/gpt-transcribe had an average word error rate of 5.49 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), openai/whisper-large-v3 had an average word error rate of 6.66 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), openai/whisper-large-v3-turbo had an average word error rate of 7.17 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), OpenMOSS-Team/MOSS-Transcribe-Diarize had an average word error rate of 5.54 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), OtterAI/speech-v3.1 had an average word error rate of 3.88 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), Qwen/Qwen3-ASR-0.6B-hf had an average word error rate of 5.77 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), Qwen/Qwen3-ASR-1.7B-hf had an average word error rate of 4.94 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), reson8/resonant-1 had an average word error rate of 4.64 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), smallestai/pulse had an average word error rate of 5.08 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), sophea/asr-k1 (preview) had an average word error rate of 4.61 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), soundsgoodai/Zipformer-cr-ctc-transducer-XL-290M (fast-gpu-asr, ctc_greedy_search) had an average word error rate of 7.64 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), soundsgoodai/Zipformer-cr-ctc-transducer-XL-290M (fast-gpu-asr, transducer_modified_beam_search) had an average word error rate of 6.22 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), sprag/symphony had an average word error rate of 5.27 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), TheStageAI/thewhisper-large-v3-turbo had an average word error rate of 5.24 percent across the leaderboard's 9 public English test sets.
- On the Hugging Face Open ASR Leaderboard (its version of October 17, 2026, read on October 20, 2026), zoom/scribe_v2_pro had an average word error rate of 4.08 percent across the leaderboard's 9 public English test sets.