Open source automatic speech recognition model known for robustness across accents and noise; often used to transcribe crypto podcasts and dev calls.