In the Voxtral paper's Table 3, ElevenLabs Scribe had a word error rate of 1.80 percent on LibriSpeech test-clean.

Quote found word for word on the live source, last checked October 19, 2026

In the source's own words

“Model LS-C LS-O GS VP SB C-4 SPGI E21 10m E22 10m ... ElevenLabs Scribe 1.80 3.44 10.52 6.95 10.62 8.35 3.16 7.39 9.16”

Voxtral, Mistral AI, arXiv (Audio and Speech Processing), 2025, Table 3 (short-form English), LibriSpeech test-clean column.

What it measures: Table 3 of the paper gives word error rate for each English test set. The paper calls the Voxtral results macro-averages in its figures, and this row is one test set.
What it does not show: One paper's runs under its own text normalization and decoding, so compare systems within this table, not with other papers' numbers. The paper gives no date per row. The paper names this system "ElevenLabs Scribe" in Table 3 and "Scribe" in Tables 4 to 6; this row treats them as one system.

A row of Speech Recognition Accuracy: Word Error Rates, as ElevenLabs Scribe, LS test-clean: 1.8 percent.