Tasks and inputs
VibeVoice ASR Streaming 1.5B is listed for Automatic speech recognition, Speaker Diarization. It accepts audio and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
SyncDev model directory · Profile updated Oct 8, 2026
VibeVoice ASR Streaming 1.5B transcribes incoming speech with speaker attribution, custom hotwords and support for ten languages.
The specification score summarizes recorded features and limits. It is not a benchmark result.
Recorded modality support shows what this model can accept and produce. An undocumented modality is shown as unknown rather than assumed to be unsupported.
Accepts
Produces
VibeVoice ASR Streaming 1.5B transcribes incoming speech with speaker attribution, custom hotwords and support for ten languages.
This checkpoint identifies who said what as audio arrives. It supports Chinese, English, French, German, Italian, Japanese, Korean, Portuguese, Russian and Spanish. Custom hotwords can supply names and technical vocabulary that matter to the application.
The 1.5B label identifies the language-model backbone; the complete speech checkpoint also includes audio components. It is a useful first candidate when memory and responsiveness matter, but measure the full model footprint rather than assuming all weights total 1.5B parameters.
Use several speakers, an interruption and a term included in the hotword list. Check the transcript and speaker labels against the original recording. Treat partial recognition as provisional until the stream finalizes it. Measure words missed, substitutions and speaker confusion separately; a transcript can be lexically accurate while assigning the wrong person.
Follow the VibeVoice repository’s streaming inference instructions and keep audio chunking consistent across runs. The weights use the MIT licence. The model produces transcription text rather than synthesized speech, and no universal hosted price is established by this checkpoint.
Compare the model on the work you need to complete.
VibeVoice ASR Streaming 1.5B is listed for Automatic speech recognition, Speaker Diarization. It accepts audio and produces text. Use the publisher’s input format and preprocessing for your chosen deployment.
Measure a completed task using the prompts, documents and tools your application will use. Include reasoning and retries in the result.
The weights are downloadable under MIT. Check runtime support and size the complete model and working context for your expected concurrency.
These records share a directory category. Check the source card, output type and deployment requirements before treating a model as a replacement.
Gemini 3.5 Transcribe Live from Google: Speech transcription model for accurate audio-to-text and captioning workflows.
xAI
Grok Voice STT 1.0 from xAI: Grok Voice STT 1.0 is xAI's speech-to-text model. It supports transcription with word-level timestamps, optional speaker diarization, and multichannel audio.
OpenAI
GPT Realtime Whisper from OpenAI: Streaming speech-to-text model for low-latency transcript deltas from live audio.
openai
openai/whisper-large-v3-turbo is a source-linked Hugging Face repository assigned to Automatic speech recognition. Its recorded task interface accepts audio and produces text.
OpenAI
Whisper 3 Large from OpenAI: Open Whisper checkpoint for robust multilingual transcription and captioning.
OpenAI
Whisper Large v3 Turbo from OpenAI: Speech transcription model for accurate audio-to-text and captioning workflows.