# Jais-ASR Arabic speech evaluation: Voice Benchmark Profile

> IFM reports 22.85% word error rate and 8.37% character error rate, macro-averaged over six Arabic speech test sets. Lower error rates are better.

- Scope: Six public Arabic ASR test sets; separate Mixat code-switched test
- Measurement lanes: Voice experience
- Primary metric: Word error rate and character error rate (%, lower is better)
- Available evidence: Exact provider tables in the model card and October 2026 launch blog
- Owner: IFM
- Source snapshots refreshed: 2026-10-15

- Provider report retrieved: 2026-10-15

## Arabic transcription accuracy

IFM-reported WER and CER (%). Lower is better. The published average is a macro-average across the six test sets.

| Test set | WER (%) | CER (%) |
| --- | --- | --- |
| [Casablanca (8 dialects)](/models/jais-asr) | 46.56 | 18.02 |
| [Common Voice 18 (MSA)](/models/jais-asr) | 8.56 | 2.70 |
| [MASC clean](/models/jais-asr) | 16.12 | 3.76 |
| [MASC noisy](/models/jais-asr) | 25.44 | 7.51 |
| [MGB-2](/models/jais-asr) | 9.59 | 3.92 |
| [SADA](/models/jais-asr) | 30.85 | 14.33 |
| [Macro-average (six sets)](/models/jais-asr) | 22.85 | 8.37 |

## Emirati code-switched speech (Mixat)

Separate 1,585-clip Mixat test set; Jais-ASR in code-switch mode.

| Model | WER (%) | CER (%) |
| --- | --- | --- |
| [Jais-ASR](/models/jais-asr) | 17.8 | 6.0 |

## Interpretation limit

IFM evaluated Jais-ASR using the public leaderboard script. These are provider-reported measurements, not independently verified leaderboard submissions. The six-set Arabic macro-average, Mixat, and internal Arabic Speech Benchmark have different samples and aggregation rules and are not pooled. Speech error rates stay separate from weighted text-model scores. October 15 is the source retrieval date, not an established evaluation or first-release day.

## Primary sources

- [Owner page](https://ifm.ai/blog/jais-asr/)
- [Jais-ASR model card](https://huggingface.co/IFM/Jais-ASR)
- [Open Universal Arabic ASR Leaderboard protocol](https://github.com/Natural-Language-Processing-Elm/open_universal_arabic_asr_leaderboard)

Canonical page: https://benchlm.ai/voice-benchmarks/jais-asr-arabic-evaluation