List your company
ResearchFormerly WizWand

Automatic Speech Recognition on Earnings 22

9.32WERQwen3-ASR-1.7B
-11.1448126.9926265.13403.2674Feb 20, 2025Apr 24, 2025Jun 26, 2025Aug 28, 2025Oct 30, 2025Jan 1, 2026Mar 6, 2026
Updated 4mo ago

Evaluation results

MethodLinks
Qwen3-ASR-1.7BDecoding Architecture=AR, Params (B)=2.02025.089.3220.125.62
Whisfusion + rerankDecoding Architecture=NAR, Params (B)=2.42025.089.64143.276.07
Phi-4-multimodalArchitecture Type=AR, Params (B)=5.6, RTFx=4.85, Avg. WER=6.72026.0210.2——
WhisfusionDecoding Architecture=NAR, Params (B)=1.52025.0810.33173.456.55
FastWhisper-largeSize/h=1,634, Params/B=0.74, Evaluation Protocol=short-form2026.0110.5——
Whisper-large-v3Decoding Architecture=AR, Params (B)=1.62025.0810.6435.337.11
Voxtral-MiniArchitecture Type=AR, Params (B)=3.0, RTFx=12.19, Avg. WER=8.02026.0210.7——
MDM-ASRArchitecture Type=NAR, Params (B)=1.0, RTFx=46.81, Avg. WER=6.92026.0210.7——
Whisper-turboDecoding Architecture=AR, Params (B)=0.82025.0810.88143.998.86
Whisper-large-v3Architecture Type=AR, Params (B)=1.5, RTFx=12.83, Avg. WER=8.52026.0211.3——
Canary-1b-v2Decoding Architecture=AR, Params (B)=1.02025.0811.3647.876.4
Whisper-large-v3Size/h=680,000, Params/B=1.50, Evaluation Protocol=short-form2026.0111.4——
Whisper-large-v3-turbo# Total Params=1.5B2025.1111.63——
distil-Whisper-large-v3Size/h=22,000, Params/B=0.75, Evaluation Protocol=short-form2026.0111.8——
CanarySize/h=31,200, Params/B=1.00, Evaluation Protocol=short-form2026.0112.2——
OWSM-CTC v3.1Decoding Architecture=NAR, Params (B)=1.02025.0812.71665.748.27
Canary-1b-flashArchitecture Type=AR, Params (B)=1.0, RTFx=29.25, Avg. WER=7.22026.0212.8——
CrisperWhisperSize/h=10,000, Params/B=1.50, Evaluation Protocol=short-form2026.0112.9——
Parakeet-CTCArchitecture Type=NAR, Params (B)=1.1, RTFx=120.20, Avg. WER=8.32026.0213.8——
Drax (MBR, 8/16)Decoding Architecture=NAR, Params (B)=1.22025.0814.0933.168.34
Qwen2-AudioArchitecture Type=AR, Params (B)=8.4, RTFx=4.18, Avg. WER=8.32026.0214.1——
Qwen2.5-Omni-3B# Total Params=5B2025.1114.81——
DraxArchitecture Type=NAR, Params (B)=1.2, RTFx=11.32, Avg. WER=9.22026.0215.2——
OWSM-v3.1Architecture Type=AR, Params (B)=1.0, RTFx=15.52, Avg. WER=10.32026.0215.4——
Whisper-LLaDADecoding Architecture=NAR, Params (B)=8.72025.0816.117.929.82
Whisper-CDModel=Large-v3-Turbo2026.0316.16——
OWSM-CTCArchitecture Type=NAR, Params (B)=1.0, RTFx=132.54, Avg. WER=11.42026.0217.2——
LFM2-Audio-1.5B# Total Params=1.5B2025.1119.75——
MMS-allDecoding Architecture=NAR, Params (B)=1.02025.0822.78681.2112.93
Whisper BaselineModel=Large-v3-Turbo2026.0333.25——
ChunkFormercontext size [latt, c, r]=[128, 64, 128]2025.0248.56——
ChunkFormercontext size [latt, c, r]=[256, 128, 128]2025.0248.67——
ChunkFormercontext size [latt, c, r]=[128, 256, 128]2025.0248.68——
Efficient Conformerdecoding_mode=full-context2025.0252.53——
Squeezeformerdecoding_mode=full-context2025.0253.44——
ChunkFormerdecoding_mode=full-context2025.0254.55——
Conformerdecoding_mode=full-context2025.0254.91——
Whisper-CDModel=Large-v32026.0357.08——
Whisper BaselineModel=Large-v32026.03520.94——