Fact · Speech recognition
2.1percent word error rate
The Conformer model reached word error rates of 2.1 percent on LibriSpeech test-clean and 4.3 percent on test-other without a language model, and 1.9 and 3.9 percent with one.
Checked against its source on October 16, 2026
In the source's own words
“our model achieves WER of 2.1%/4.3% without using a language model and 1.9%/3.9% with an external language model on test/testother”
Conformer: Convolution-augmented Transformer for Speech Recognition, Anmol Gulati et al., arXiv (Interspeech 2020), 2020, Abstract.
Open the original- What it measures
- Conformer combines convolution, which captures local sound detail, with transformer self-attention, which captures long-range context.
- What it does not show
- Read audiobook speech in 2020; later systems score lower on this benchmark.
Source
Cited on
- Speech Recognition: What It Is and How Well It WorksTopic
- How Speech Recognition Works, From Sound to TextGuide
- Is Speech to Text AI? What Kind of AI Voice Typing UsesGuide
- Word Error Rate (WER): How to Calculate and Read ItGuide
- Speech Recognition Models: How Neural ASR WorksGuide
- Speech Recognition Benchmarks and LeaderboardsGuide
- Which neural network is best for speech recognition?Question
- ConformerGlossary