One 600M model replaces 40 language checkpoints without swapping
nvidia's nemotron 3.5 asr supports 40 languages in a single 600m-parameter checkpoint with native punctuation and low-latency streaming via cache-aware fastconformer.
nvidia introduced nemotron 3.5 asr, a 600m-parameter speech model designed to handle 40 languages and dialects within a single unified checkpoint, removing the need for per-language deployments or separate punctuation models. the architecture utilizes a cache-aware fastconformer encoder to process audio frames exactly once, significantly reducing latency and compute costs while maintaining accuracy compared to traditional buffered streaming approaches. it features native output of punctuated and capitalized text, and offers optional language id conditioning for either specified or automatic language detection in mixed-language scenarios.