Nemotron 3.5 Asr Streaming 0.6b

Nemotron 3.5 ASR Streaming 0.6B GGUF is a quantized version of NVIDIA's Nemotron 3.5 automatic speech recognition (ASR) streaming model, packaged in the GGUF format for use with the Vibe app. The model features a 0.6 billion parameter architecture and is designed for automatic speech recognition tasks. According to the evidence, Vibe utilizes this model in conjunction with a Silero voice activity detection (VAD) model to enable bounded, long-form transcription. The quantized build provided is specifically identified as Q4_K_M, with a model size of 496 MB and 4-bit quantization. The model is intended for users who require ASR capabilities within the Vibe app ecosystem, particularly for scenarios involving long-form or continuous audio transcription. The evidence also notes that the upstream base model is nvidia/nemotron-3.5-asr-streaming-0.6b, and users are directed to review its model card and OpenMDW 1.1 license before use. However, there is no further detail about licensing terms or pricing in the provided evidence. The tool is not currently deployed by any inference provider as per the available information. Delivery of the model is via the GGUF format, which is compatible with the Vibe app and potentially other systems that support this format. The evidence does not elaborate on standalone usage, supported platforms beyond Vibe, or integration options outside the stated context. There is no information about additional features, supported languages, or user roles beyond its use for ASR within the Vibe app. Overall, Nemotron 3.5 ASR Streaming 0.6B GGUF is positioned as a quantized ASR model for streaming transcription, primarily intended for use by the Vibe app in combination with a VAD model for handling long-form audio. Further technical details or broader applicability are not addressed in the provided evidence.