HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks

VibeVoice-Realtime-0.5B

microsoft/VibeVoice-Realtime-0.5B

kind
original
quant
bf16 (16-bit)
family
misc
modality
text-to-speech
root model
Qwen/Qwen2.5-0.5B
provider
microsoft
license
mit
language
en
context
—
library
transformers
vRAM (GB)
—
size on disk
—
downloads
661,602
likes
1,259
created
—
last modified
2025-12-12T15:35:46+00:00 (2025)
huggingface
https://huggingface.co/microsoft/VibeVoice-Realtime-0.5B
tags
transformerssafetensorsvibevoice_streamingRealtime TTSStreaming text inputLong-form speech generationtext-to-speechenarxiv:2508.19205arxiv:2412.08635base_model:Qwen/Qwen2.5-0.5Bbase_model:finetune:Qwen/Qwen2.5-0.5Blicense:mitendpoints_compatibleregion:us

Benchmark scores (2)

benchmarkcategoryscorecontextconfevidence
Speaker SIMspeech-tts64.3low| VALL-E 2 | 2.40 | 0.643 |
WERspeech-asr2.4low| VALL-E 2 | 2.40 | 0.643 |

Discussions (0)

none scraped.

Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.