Voice of India - Urdu: leaderboard
Speech recognition on the Urdu part of Voice of India (2026), a closed benchmark of unscripted telephone conversations in Indian languages, scored with an orthographically informed word error rate that accepts permissible spelling variants. Percent, can exceed 100; lower is better.
Metric: Orthographically informed word error rate (OIWER, %; can exceed 100) on the Urdu unscripted telephone-conversation utterances of the closed Voice of India benchmark, transcripts with lattices of permissible spelling variants; systems run at default inference settings and only where they support the language; lower is better. Source: arxiv.org. Saturation forecast: Around March 2028. 11 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Sarvam Audio | 7 | |
| 2 | IndicConformer | 8.1 | |
| 3 | Gemini 3 Pro | 9.1 | #145 |
| 4 | Gemini 3 Flash | 11.9 | #175 |
| 5 | omniASR_LLM_7B | 16 | |
| 6 | omniASR_LLM_1B | 17.2 | |
| 7 | Microsoft Speech-to-Text | 25.2 | |
| 8 | elevenlabs-scribe_v2 | 25.5 | |
| 9 | AssemblyAI Universal | 31.9 | |
| 10 | GPT-4o Transcribe | 35.4 | |
| 11 | gpt-4o-mini-transcribe | 52 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=voice-of-india-urdu · How It Works · Data refreshed daily, snapshot 2026-11-21.