January 26, 2026
Announcing WER Testing and AWS Transcribe Support
Two new features are now available: Word Error Rate (WER) testing and AWS Transcribe as a provider. Both are in early stages and we'd love your feedback.
WER Testing Tool
We've added a new WER testing page where you can upload your own audio along with a reference transcript to calculate Word Error Rate for different STT providers. This helps you measure transcription accuracy on your specific content.
This feature is still under development and may have rough edges. We're looking for feedback on:
- Usability of the interface
- Accuracy of WER calculations
- Supported audio formats and limitations
- Additional metrics you'd like to see
AWS Transcribe Support
AWS Transcribe is now available as a provider option. You can test it alongside other providers to compare performance and accuracy on your audio files.
Try it out now on our AWS Transcribe testing page.
This integration is also in its early stages - no native config yet.
More Providers Coming Soon
We're actively working on adding more speech-to-text providers to give you even more options for comparison. Stay tuned for updates!