Fact · Speech recognition
2.3percent AA-WER
In the Artificial Analysis AA-WER v2.0 benchmark, ElevenLabs Scribe v2 had the lowest error rate at 2.3 percent, ahead of Google's Gemini 3 Pro at 2.9 percent.
Checked against its source on October 17, 2026
In the source's own words
“ElevenLabs Scribe v2 leads at 2.3% AA-WER, followed by Google's Gemini 3 Pro at 2.9%”
AA-WER v2.0: Speech to Text Accuracy Benchmark, Artificial Analysis, 2026, Key results.
Open the original- What it measures
- A single independent benchmark with a published method. The score is a weighted average of three datasets, and the leaderboard shows 2.2 percent for the same model in a later snapshot.
- What it does not show
- It does not test Whisper in the article's headline table, so it is not a direct Scribe against Whisper comparison.