As the title says, I'm looking for the best speech-to-text models to infer on indian languages, both closed and open source models I've heard sarvam released their proprietary "saaras v3" but how good is it ? any open source alternatives? ( I'd prefer getting started with the model right away than t
