OpenCodePapers

speech-recognition-on-ami-cleaned

Speech Recognition
Dataset Link
Results over time
Click legend items to toggle metrics. Hover points for model names.
Leaderboard
Click a sortable column header to sort.
PaperCodeWord Error Rate (WER)ModelNameReleaseDate
Introducing Zoom AI Services6.94zoom/scribe_v12026-03-10
Azure Speech at Build 2026: Powering Voice Agents with Real-Time and Life-like Experiences6.95microsoft/azure-speech-06-20262026-06-04
NLE: Non-autoregressive LLM-based ASR by Transcript Editing✓ Link6.96ibm-granite/granite-speech-4.1-2b-nar2026-04-01
Introducing Cohere-transcribe: state-of-the-art speech recognition7.01CohereLabs/cohere-transcribe-03-20262026-03-26
✓ Link7.06ibm-granite/granite-speech-4.1-2b2026-04-29
CrisperWhisper: Accurate Timestamps on Verbatim Speech Transcriptions✓ Link7.1nyrahealth/CrisperWhisper2024-08-29
Boson AI Launches Higgs STT 3 Speech-to-Text Model✓ Link7.19bosonai/higgs-audio-v3-stt2026-03-18
✓ Link7.37ibm-granite/granite-4.0-1b-speech2026-03-06
Granite-speech: open-source speech-aware LLMs with strong English ASR capabilities✓ Link7.7ibm-granite/granite-speech-3.3-8b2025-06-19
Granite-speech: open-source speech-aware LLMs with strong English ASR capabilities✓ Link7.8ibm-granite/granite-speech-3.3-2b2025-06-19
7.8OpenMOSS-Team/MOSS-Transcribe-preview-2B2026-06-26
Data-Efficient On-Policy Distillation for Automatic Speech Recognition✓ Link7.89AutoArk-AI/ARK-ASR-3B2026-06-22
✓ Link7.91nvidia/canary-qwen-2.5b2025-07-17
Introducing Scribe v28.2elevenlabs/scribe_v22026-01-09
Introducing Resonant-1 and Resonant-1-flash8.21reson8/resonant-1-flash2026-04-08
Introducing Resonant-1 and Resonant-1-flash8.22reson8/resonant-12026-04-08
Qwen3-ASR Technical Report✓ Link8.31Qwen/Qwen3-ASR-1.7B2026-01-28
Boson AI Launches Higgs STT 3 Speech-to-Text Model✓ Link8.37bosonai/higgs-audio-v3-8b-stt-v22026-04-27
Data-Efficient On-Policy Distillation for Automatic Speech Recognition✓ Link8.6AutoArk-AI/ARK-ASR-0.6B2026-05-25
Stateful Conformer with Cache-based Inference for Streaming Automatic Speech Recognition✓ Link8.84nvidia/nemotron-speech-streaming-en-0.6b2026-03-13
Moonshine v2: Ergodic Streaming Encoder ASR for Latency-Critical Speech Applications✓ Link8.93usefulsensors/moonshine-streaming-medium2026-02-12
Efficient Sequence Transduction by Jointly Predicting Tokens and Durations✓ Link9.1nvidia/parakeet-tdt-0.6b-v22025-05-01
Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs9.11microsoft/Phi-4-multimodal-instruct2025-02-24
Pulse STT: Fast, Accurate Speech-to-Text9.12smallestai/pulse2026-01-28
Canary-1B-v2 & Parakeet-TDT-0.6B-v3: Efficient and High-Performance Models for Multilingual ASR and AST✓ Link9.41nvidia/parakeet-tdt-0.6b-v32025-08-14
Qwen3-ASR Technical Report✓ Link9.47Qwen/Qwen3-ASR-0.6B2026-01-28
Avalon: ASR for Human–AI Interaction9.71aquavoice/avalon-v1-en2025-08-22
Ursa 2: Elevating speech recognition across 50+ languages9.96speechmatics/enhanced2024-10-11
Streaming Sequence-to-Sequence Learning with Delayed Streams Modeling✓ Link10.57kyutai/stt-2.6b-en2025-06-06
Training and Inference Efficiency of Encoder-Decoder Speech Models✓ Link10.58nvidia/canary-1b-flash2025-03-07
Moonshine v2: Ergodic Streaming Encoder ASR for Latency-Critical Speech Applications✓ Link10.79usefulsensors/moonshine-streaming-small2026-02-12
Less is More: Accurate Speech Recognition & Translation without Web-Scale Data✓ Link10.87nvidia/canary-1b2024-02-07
Introducing Universal-3 Pro: A new class of speech language model optimized for Voice AI11.06assemblyai/universal-3-pro2026-02-03
OWSM v4: Improving Open Whisper-Style Speech Models via Data Scaling and Cleaning✓ Link11.11espnet/owsm_ctc_v4_1B2025-01-16
VibeVoice-ASR Technical Report✓ Link12.03microsoft/VibeVoice-ASR-HF2026-03-02
Robust Knowledge Distillation via Large-Scale Pseudo Labelling✓ Link12.09distil-whisper/distil-large-v3.52024-12-05
Training and Inference Efficiency of Encoder-Decoder Speech Models✓ Link12.1nvidia/canary-180m-flash2025-03-11
Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition✓ Link12.24nvidia/parakeet-ctc-1.1b2023-12-28
Efficient Sequence Transduction by Jointly Predicting Tokens and Durations✓ Link12.69nvidia/parakeet-tdt_ctc-110m2024-09-17
Zipformer: A faster and better encoder for automatic speech recognition✓ Link12.8soundsgoodai/Zipformer-transducer-XL-290M2026-05-12
Efficient Sequence Transduction by Jointly Predicting Tokens and Durations✓ Link12.87nvidia/parakeet-tdt-1.1b2024-01-25
Canary-1B-v2 & Parakeet-TDT-0.6B-v3: Efficient and High-Performance Models for Multilingual ASR and AST✓ Link13.02nvidia/canary-1b-v22025-08-14
Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition✓ Link13.1nvidia/parakeet-ctc-0.6b2023-12-28
Voxtral✓ Link13.18mistralai/Voxtral-Small-24B-25072025-07-15
Conformer: Convolution-augmented Transformer for Speech Recognition✓ Link13.2nvidia/stt_en_conformer_ctc_large2022-04-09
Voxtral Realtime✓ Link13.34mistralai/Voxtral-Mini-4B-Realtime-26022026-02-11
OWSM-CTC: An Open Encoder-Only Speech Foundation Model for Speech Recognition, Translation, and Language Identification✓ Link13.35espnet/owsm_ctc_v3.1_1B2024-02-23
Stateful Conformer with Cache-based Inference for Streaming Automatic Speech Recognition✓ Link13.42nvidia/nemotron-3.5-asr-streaming-0.6b2026-06-04
Voxtral✓ Link13.56mistralai/Voxtral-Mini-3B-25072025-07-01
Robust Speech Recognition via Large-Scale Weak Supervision✓ Link13.63openai/whisper-large-v32023-11-06
Moonshine: Speech Recognition for Live Transcription and Voice Commands✓ Link13.68usefulsensors/moonshine-base2024-10-21
LiteASR: Efficient Automatic Speech Recognition with Low-Rank Approximation✓ Link13.69efficient-speech/lite-whisper-large-v3-acc2025-02-26
Niagara-38m Sets a New Benchmark for Edge Speech Recognition13.76abr-ai/niagara-38m-batch.en2026-04-15
Omnilingual ASR: Open-Source Multilingual Speech Recognition for 1600+ Languages✓ Link13.86facebook/omniASR-LLM-7B-v22025-12-12
Robust Speech Recognition via Large-Scale Weak Supervision✓ Link13.87openai/whisper-large-v3-turbo2024-10-01
On the Effects of Heterogeneous Data Sources on Speech-to-Text Foundation Models✓ Link13.88espnet/owsm_ctc_v3.2_ft_1B2024-09-24
GLM-ASR-Nano: A robust, open-source speech recognition model✓ Link13.95zai-org/GLM-ASR-Nano-25122025-12-09
Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition✓ Link14.02nvidia/parakeet-rnnt-1.1b2023-12-27
LiteASR: Efficient Automatic Speech Recognition with Low-Rank Approximation✓ Link14.08efficient-speech/lite-whisper-large-v3-turbo-acc2025-02-26
LiteASR: Efficient Automatic Speech Recognition with Low-Rank Approximation✓ Link14.14efficient-speech/lite-whisper-large-v32025-02-26
Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition✓ Link14.34nvidia/parakeet-rnnt-0.6b2023-12-28
Loquacious Set: 25,000 Hours of Transcribed and Diverse English Speech Recognition Data for Research and Commercial Use✓ Link15.57speechbrain/asr-conformer-largescaleasr2025-02-06
Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition✓ Link16.12nvidia/stt_en_fastconformer_ctc_large2023-06-08
Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition✓ Link16.35nvidia/stt_en_fastconformer_transducer_large2023-06-08
LiteASR: Efficient Automatic Speech Recognition with Low-Rank Approximation✓ Link16.53efficient-speech/lite-whisper-large-v3-fast2025-02-26
Niagara-38m Sets a New Benchmark for Edge Speech Recognition16.71abr-ai/niagara-19m-batch.en2026-04-15
Moonshine v2: Ergodic Streaming Encoder ASR for Latency-Critical Speech Applications✓ Link16.94usefulsensors/moonshine-streaming-tiny2026-02-12
Conformer: Convolution-augmented Transformer for Speech Recognition✓ Link17.19nvidia/stt_en_conformer_transducer_small2022-06-01
Omnilingual ASR: Open-Source Multilingual Speech Recognition for 1600+ Languages✓ Link17.44facebook/omniASR-CTC-7B-v22025-12-12
Conformer: Convolution-augmented Transformer for Speech Recognition✓ Link17.8nvidia/stt_en_conformer_ctc_small2023-06-12
Moonshine: Speech Recognition for Live Transcription and Voice Commands✓ Link18.83usefulsensors/moonshine-tiny2024-10-21
wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations✓ Link25.83facebook/wav2vec2-large-960h-lv60-self2020-06-20
Robust wav2vec 2.0: Analyzing Domain Shift in Self-Supervised Pre-Training✓ Link27.04facebook/wav2vec2-large-robust-ft-libri-960h2021-04-02
SpeechBrain: A General-Purpose Speech Toolkit✓ Link27.1speechbrain/asr-wav2vec2-librispeech2022-06-05
HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units✓ Link28.15facebook/hubert-xlarge-ls960-ft2022-03-02
data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language✓ Link28.8facebook/data2vec-audio-large-960h2022-04-02
HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units✓ Link28.9facebook/hubert-large-ls960-ft2022-03-02
fairseq S2T: Fast Speech-to-Text Modeling with fairseq✓ Link31.23facebook/wav2vec2-conformer-rope-large-960h-ft2022-04-18
Scaling Speech Technology to 1,000+ Languages✓ Link31.24facebook/mms-1b-all2023-05-27
fairseq S2T: Fast Speech-to-Text Modeling with fairseq✓ Link31.38facebook/wav2vec2-conformer-rel-pos-large-960h-ft2022-04-18
wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations✓ Link32.13facebook/wav2vec2-large-960h2020-06-20
wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations✓ Link35.41facebook/wav2vec2-base-960h2022-03-02
data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Language✓ Link37.29facebook/data2vec-audio-base-960h2022-03-02
Check out how to contribute new results. Then edit this file.