{ "updated": "2026-04-28", "dataset": { "name": "PriMock57", "files": 55, "total_words": 81236, "excluded_files": [ "day1_consultation07", "day3_consultation03" ] }, "models": [ { "rank": 1, "name": "google-gemini-3-pro-preview", "wer": 0.0835, "accuracy": 0.9165, "avg_speed_sec": 64.5, "best_wer": 0.0417, "worst_wer": 0.1655, "wer_std": 0.0268, "files_evaluated": 54, "m_wer": 0.0137, "drug_m_wer": 0.0109, "note": "*54/55 files evaluated" }, { "rank": 2, "name": "google-gemini-2.5-pro", "wer": 0.0815, "accuracy": 0.9185, "avg_speed_sec": 56.4, "best_wer": 0.0368, "worst_wer": 0.1434, "wer_std": 0.0238, "files_evaluated": 55, "m_wer": 0.0152, "drug_m_wer": 0.0193 }, { "rank": 3, "name": "vibevoice-asr-9b", "wer": 0.0834, "accuracy": 0.9166, "avg_speed_sec": 96.7, "best_wer": 0.0391, "worst_wer": 0.1796, "wer_std": 0.028, "files_evaluated": 55, "m_wer": 0.0181, "drug_m_wer": 0.0451 }, { "rank": 4, "name": "google-gemini-3-flash-preview", "wer": 0.1133, "accuracy": 0.8867, "avg_speed_sec": 51.5, "best_wer": 0.0563, "worst_wer": 0.2183, "wer_std": 0.0376, "files_evaluated": 55, "m_wer": 0.0203, "drug_m_wer": 0.03 }, { "rank": 5, "name": "soniox-stt-async-v4", "wer": 0.0918, "accuracy": 0.9082, "avg_speed_sec": 46.2, "best_wer": 0.0401, "worst_wer": 0.1631, "wer_std": 0.0296, "files_evaluated": 55, "m_wer": 0.0206, "drug_m_wer": 0.0536 }, { "rank": 6, "name": "elevenlabs-scribe_v2", "wer": 0.0972, "accuracy": 0.9028, "avg_speed_sec": 43.5, "best_wer": 0.0381, "worst_wer": 0.6569, "wer_std": 0.0816, "files_evaluated": 55, "m_wer": 0.0254, "drug_m_wer": 0.0279 }, { "rank": 7, "name": "assemblyai-universal-3-pro-medical", "wer": 0.0955, "accuracy": 0.9045, "avg_speed_sec": 37.3, "best_wer": 0.052, "worst_wer": 0.1568, "wer_std": 0.0251, "files_evaluated": 55, "m_wer": 0.0283, "drug_m_wer": 0.0494 }, { "rank": 8, "name": "qwen3-asr-1.7b", "wer": 0.09, "accuracy": 0.91, "avg_speed_sec": 6.8, "best_wer": 0.0381, "worst_wer": 0.4413, "wer_std": 0.0532, "files_evaluated": 55, "m_wer": 0.0314, "drug_m_wer": 0.0708 }, { "rank": 9, "name": "deepgram-nova-3-medical", "wer": 0.0905, "accuracy": 0.9095, "avg_speed_sec": 12.9, "best_wer": 0.0479, "worst_wer": 0.1406, "wer_std": 0.0242, "files_evaluated": 55, "m_wer": 0.0317, "drug_m_wer": 0.0794 }, { "rank": 10, "name": "mai-transcribe-1", "wer": 0.1152, "accuracy": 0.8848, "avg_speed_sec": 21.8, "best_wer": 0.0635, "worst_wer": 0.1783, "wer_std": 0.0283, "files_evaluated": 55, "m_wer": 0.0333, "drug_m_wer": 0.088 }, { "rank": 11, "name": "elevenlabs-scribe_v1", "wer": 0.1087, "accuracy": 0.8913, "avg_speed_sec": 36.3, "best_wer": 0.0459, "worst_wer": 0.6688, "wer_std": 0.0887, "files_evaluated": 55, "m_wer": 0.0359, "drug_m_wer": 0.0601 }, { "rank": 12, "name": "google-gemini-2.5-flash", "wer": 0.0945, "accuracy": 0.9055, "avg_speed_sec": 20.2, "best_wer": 0.0401, "worst_wer": 0.3564, "wer_std": 0.0447, "files_evaluated": 55, "m_wer": 0.0365, "drug_m_wer": 0.0815 }, { "rank": 13, "name": "openai-gpt-4o-mini-transcribe-2025-12-15", "wer": 0.1118, "accuracy": 0.8882, "avg_speed_sec": 40.4, "best_wer": 0.0616, "worst_wer": 0.2325, "wer_std": 0.0324, "files_evaluated": 55, "m_wer": 0.0368, "drug_m_wer": 0.088 }, { "rank": 14, "name": "parakeet-tdt-1.1b", "wer": 0.0903, "accuracy": 0.9097, "avg_speed_sec": 12.3, "best_wer": 0.0526, "worst_wer": 0.1474, "wer_std": 0.0228, "files_evaluated": 55, "m_wer": 0.0368, "drug_m_wer": 0.1373 }, { "rank": 15, "name": "voxtral-mini-transcribe-v1-chat", "wer": 0.1185, "accuracy": 0.8815, "avg_speed_sec": 22.4, "best_wer": 0.0528, "worst_wer": 0.4689, "wer_std": 0.057, "files_evaluated": 55, "m_wer": 0.0403, "drug_m_wer": 0.0966 }, { "rank": 16, "name": "voxtral-mini-transcribe-v2", "wer": 0.1164, "accuracy": 0.8836, "avg_speed_sec": 18.4, "best_wer": 0.0528, "worst_wer": 0.2364, "wer_std": 0.0345, "files_evaluated": 55, "m_wer": 0.041, "drug_m_wer": 0.1094 }, { "rank": 17, "name": "voxtral-mini-4b-realtime", "wer": 0.1189, "accuracy": 0.8811, "avg_speed_sec": 133.9, "best_wer": 0.0499, "worst_wer": 0.2104, "wer_std": 0.0346, "files_evaluated": 55, "m_wer": 0.041, "drug_m_wer": 0.103 }, { "rank": 18, "name": "groq-whisper-large-v3-turbo", "wer": 0.1214, "accuracy": 0.8786, "avg_speed_sec": 8.0, "best_wer": 0.0606, "worst_wer": 0.194, "wer_std": 0.0316, "files_evaluated": 55, "m_wer": 0.0432, "drug_m_wer": 0.1223 }, { "rank": 19, "name": "cohere-transcribe-03-2026", "wer": 0.1181, "accuracy": 0.8819, "avg_speed_sec": 3.9, "best_wer": 0.0587, "worst_wer": 0.2551, "wer_std": 0.0361, "files_evaluated": 55, "m_wer": 0.0432, "drug_m_wer": 0.1481 }, { "rank": 20, "name": "openai-whisper-1", "wer": 0.132, "accuracy": 0.868, "avg_speed_sec": 104.3, "best_wer": 0.056, "worst_wer": 1.0469, "wer_std": 0.1292, "files_evaluated": 55, "m_wer": 0.0451, "drug_m_wer": 0.0858 }, { "rank": 21, "name": "canary_1b_flash_lcs", "wer": 0.1203, "accuracy": 0.8797, "avg_speed_sec": 23.4, "best_wer": 0.0596, "worst_wer": 0.2046, "wer_std": 0.0333, "files_evaluated": 55, "m_wer": 0.0451, "drug_m_wer": 0.1416 }, { "rank": 22, "name": "groq-whisper-large-v3", "wer": 0.1193, "accuracy": 0.8807, "avg_speed_sec": 8.6, "best_wer": 0.0635, "worst_wer": 0.1886, "wer_std": 0.031, "files_evaluated": 55, "m_wer": 0.046, "drug_m_wer": 0.118 }, { "rank": 23, "name": "parakeet-parakeet-tdt-0.6b-v2", "wer": 0.1075, "accuracy": 0.8925, "avg_speed_sec": 5.4, "best_wer": 0.0596, "worst_wer": 0.1779, "wer_std": 0.0292, "files_evaluated": 55, "m_wer": 0.046, "drug_m_wer": 0.1567 }, { "rank": 24, "name": "mlx-community_whisper-large-v3-turbo", "wer": 0.1165, "accuracy": 0.8835, "avg_speed_sec": 12.9, "best_wer": 0.0508, "worst_wer": 0.2947, "wer_std": 0.0395, "files_evaluated": 55, "m_wer": 0.047, "drug_m_wer": 0.1223 }, { "rank": 25, "name": "whisperkit-large-v3-v20240930_turbo", "wer": 0.1228, "accuracy": 0.8772, "avg_speed_sec": 21.4, "best_wer": 0.0538, "worst_wer": 0.1956, "wer_std": 0.034, "files_evaluated": 55, "m_wer": 0.0473, "drug_m_wer": 0.1245 }, { "rank": 26, "name": "openai-gpt-4o-mini-transcribe", "wer": 0.136, "accuracy": 0.864, "avg_speed_sec": 23.2, "best_wer": 0.0596, "worst_wer": 0.3945, "wer_std": 0.0573, "files_evaluated": 55, "m_wer": 0.0486, "drug_m_wer": 0.0966 }, { "rank": 27, "name": "qwen3-asr-0.6b", "wer": 0.0983, "accuracy": 0.9017, "avg_speed_sec": 5.1, "best_wer": 0.0391, "worst_wer": 0.2039, "wer_std": 0.0299, "files_evaluated": 55, "m_wer": 0.0495, "drug_m_wer": 0.1373 }, { "rank": 28, "name": "kyutai-stt-pytorch-stt-2.6b-en", "wer": 0.112, "accuracy": 0.888, "avg_speed_sec": 148.4, "best_wer": 0.0538, "worst_wer": 0.1828, "wer_std": 0.0305, "files_evaluated": 55, "m_wer": 0.0521, "drug_m_wer": 0.1395 }, { "rank": 29, "name": "glm-asr-nano-2512", "wer": 0.1084, "accuracy": 0.8916, "avg_speed_sec": 87.7, "best_wer": 0.0596, "worst_wer": 0.1781, "wer_std": 0.0255, "files_evaluated": 55, "m_wer": 0.0575, "drug_m_wer": 0.1609 }, { "rank": 30, "name": "parakeet-parakeet-tdt-0.6b-v3", "wer": 0.0935, "accuracy": 0.9065, "avg_speed_sec": 6.3, "best_wer": 0.0381, "worst_wer": 0.1568, "wer_std": 0.0244, "files_evaluated": 55, "m_wer": 0.059, "drug_m_wer": 0.206 }, { "rank": 31, "name": "nemotron-speech-streaming-0.6b", "wer": 0.1106, "accuracy": 0.8894, "avg_speed_sec": 11.7, "best_wer": 0.043, "worst_wer": 0.1853, "wer_std": 0.0304, "files_evaluated": 55, "m_wer": 0.0705, "drug_m_wer": 0.2103 }, { "rank": 32, "name": "openai-gpt-4o-transcribe", "wer": 0.1484, "accuracy": 0.8516, "avg_speed_sec": 27.9, "best_wer": 0.0587, "worst_wer": 0.6545, "wer_std": 0.0957, "files_evaluated": 55, "m_wer": 0.0787, "drug_m_wer": 0.1352 }, { "rank": 33, "name": "gemma-4-e4b-it", "wer": 0.1569, "accuracy": 0.8431, "avg_speed_sec": 185.4, "best_wer": 0.0743, "worst_wer": 0.2293, "wer_std": 0.037, "files_evaluated": 55, "m_wer": 0.079, "drug_m_wer": 0.1245 }, { "rank": 34, "name": "nvidia_canary-qwen-2.5b", "wer": 0.1294, "accuracy": 0.8706, "avg_speed_sec": 105.4, "best_wer": 0.0577, "worst_wer": 0.6109, "wer_std": 0.0726, "files_evaluated": 55, "m_wer": 0.0825, "drug_m_wer": 0.2103 }, { "rank": 35, "name": "canary-1b-v2", "wer": 0.1432, "accuracy": 0.8568, "avg_speed_sec": 9.2, "best_wer": 0.0694, "worst_wer": 0.4257, "wer_std": 0.0597, "files_evaluated": 55, "m_wer": 0.094, "drug_m_wer": 0.1803 }, { "rank": 36, "name": "granite-speech-3.3-2b", "wer": 0.1655, "accuracy": 0.8345, "avg_speed_sec": 109.7, "best_wer": 0.0528, "worst_wer": 0.3333, "wer_std": 0.0513, "files_evaluated": 55, "m_wer": 0.1102, "drug_m_wer": 0.221 }, { "rank": 37, "name": "apple-speechanalyzer", "wer": 0.1236, "accuracy": 0.8764, "avg_speed_sec": 6.0, "best_wer": 0.0655, "worst_wer": 0.1875, "wer_std": 0.0312, "files_evaluated": 55, "m_wer": 0.1197, "drug_m_wer": 0.2532 }, { "rank": 38, "name": "gemma-4-e2b-it", "wer": 0.189, "accuracy": 0.811, "avg_speed_sec": 134.6, "best_wer": 0.0831, "worst_wer": 0.2834, "wer_std": 0.0468, "files_evaluated": 55, "m_wer": 0.1222, "drug_m_wer": 0.176 }, { "rank": 39, "name": "azure-foundry-phi4", "wer": 0.3113, "accuracy": 0.6887, "avg_speed_sec": 212.8, "best_wer": 0.0697, "worst_wer": 1.051, "wer_std": 0.2776, "files_evaluated": 55, "m_wer": 0.1419, "drug_m_wer": 0.1609 }, { "rank": 40, "name": "kyutai-stt-pytorch-stt-1b-en_fr", "wer": 0.2728, "accuracy": 0.7272, "avg_speed_sec": 79.5, "best_wer": 0.0679, "worst_wer": 2.2164, "wer_std": 0.2986, "files_evaluated": 55, "m_wer": 0.199, "drug_m_wer": 0.279 }, { "rank": 41, "name": "google-medasr", "wer": 0.5254, "accuracy": 0.4746, "avg_speed_sec": 3.9, "best_wer": 0.3427, "worst_wer": 0.7233, "wer_std": 0.0771, "files_evaluated": 55, "m_wer": 0.2616, "drug_m_wer": 0.3562 }, { "rank": 42, "name": "mms-1b-all", "wer": 0.387, "accuracy": 0.613, "avg_speed_sec": 28.6, "best_wer": 0.2385, "worst_wer": 0.523, "wer_std": 0.0634, "files_evaluated": 55, "m_wer": 0.5292, "drug_m_wer": 0.7103 } ] }