Preprint
Measuring the Accuracy of Automatic Speech Recognition Solutions (HTML version)
Korbinian Kuhn et al., arXiv, 2024
Read the originalRead on October 17, 2026
- What it is
- The HTML version of the same 2024 study, which measured eleven commercial and open speech recognition services on university lecture recordings in English and German.
- How much weight it carries
- A preprint, not yet peer-reviewed when checked. The same test as the PDF, with the same limits: lecture audio and a single test set.
Facts taken from this source
- In the 2024 study, Amazon had an average word error rate of 4.4 percent on the English lecture recordings.
“Table 4. Average Word Error Rate in percent by service provider and dataset Vendor English LibriSpeech ESL German ... Amazon 4.4 6.1 7.1 18.0” (Table 4)
- In the 2024 study, Deepgram had an average word error rate of 8.3 percent on the English lecture recordings.
“Table 4. Average Word Error Rate in percent by service provider and dataset Vendor English LibriSpeech ESL German ... Deepgram 8.3 12.8 11.4 19.3” (Table 4)
- In the 2024 study, Google had the highest mean word error rate of the eleven services, with the largest spread between recordings.
“Google has the highest mean and standard deviation.” (Section 4.1, Accuracy)
- In the 2024 study, IBM had an average word error rate of 11.2 percent on the English lecture recordings.
“Table 4. Average Word Error Rate in percent by service provider and dataset Vendor English LibriSpeech ESL German ... IBM 11.2 13.2 17.3 20.6” (Table 4)
- In the 2024 study, Microsoft had an average word error rate of 4.4 percent on the English lecture recordings.
“Table 4. Average Word Error Rate in percent by service provider and dataset Vendor English LibriSpeech ESL German ... Microsoft 4.4 5.9 6.7 10.1” (Table 4)