{
  "id": 15637926,
  "title": "MAI-Transcribe-2 Leads David AI’s 21-Language Speech Recognition Benchmark",
  "url": "https://urgent.news/2026/10/19/mai-transcribe-2-leads-david-ais-21-language-speech-recognition",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-19T18:15:30.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/alifar/mai-transcribe-2-leads-david-ais-21-language-speech-recognition-benchmark-104m"
  },
  "original_language": "en",
  "account": "<company>David AI</company>'s new <term>Dai-ASR-I18N</term> benchmark reveals that <product>MAI-Transcribe-2</product> delivered the lowest transcription error rate in 18 out of 21 languages. This evaluation used 147 hours of unscripted conversation, a more realistic setting than scripted audio. The model maintained transcription errors below 10% in 16 languages and showed robustness when mixed speaker audio was combined into a single channel. While ElevenLabs Scribe v2 had the lowest error rate in the remaining three languages, including English, MAI-Transcribe-2 demonstrated the most consistent performance across all languages tested. Organizations processing multilingual spoken content should consider these findings, as they provide independent evidence of potential performance variations by language and audio setup. David AI has made the evaluation data and code publicly available to promote transparency and reproducibility in the field. However, buyers should still conduct controlled pilots using representative recordings to ensure a transcription tool meets their specific workflow requirements.",
  "summary": "Microsoft AI’s MAI-Transcribe-2 achieved the lowest transcription error rate in 18 of 21 languages in David AI’s new DAI-ASR-I18N public evaluation. The result matters because the benchmark uses 147 hours of unscripted conversation, a more demanding setting than clean, scripted audio and one that more closely resembles customer calls, interviews, meetings, and support recordings. According to…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}