HF Model & Benchmark Matrix

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

Breeze-ASR-25-gguf

handy-computer/Breeze-ASR-25-gguf

kind
gguf
quant
gguf (16-bit)
family
misc
modality
automatic-speech-recognition
root model
MediaTek-Research/Breeze-ASR-25
provider
handy-computer
license
apache-2.0
language
zh, en
context
—
library
transcribe.cpp
vRAM (GB)
—
size on disk
—
downloads
18,272
likes
0
created
—
last modified
2026-06-29T09:52:45+00:00 (2026)
huggingface
https://huggingface.co/handy-computer/Breeze-ASR-25-gguf
tags
transcribe.cppggufasrspeech-to-textwhispermediatekmandarintraditional-chinesecode-switchingautomatic-speech-recognitionzhenarxiv:2506.11130base_model:MediaTek-Research/Breeze-ASR-25base_model:quantized:MediaTek-Research/Breeze-ASR-25license:apache-2.0region:us

Benchmark scores (3)

benchmarkcategoryscorecontextconfevidence
WER Common Voicespeech-asr9.84CommonVoice16-zh-TWmedium| CommonVoice16-zh-TW | Mandarin | 9.84 | 8.95 | 11.86 | **7.97** (-19%) |
WER FLEURSspeech-asr8.12low| BF16 | [Breeze-ASR-25-BF16.gguf](https://huggingface.co/handy-computer/Breeze-ASR-25-gguf/resolve/main/Breeze-ASR-25-BF16.gguf) | 3.10 GB | 2.29% | 8.12% |
WERspeech-asr2.29low| BF16 | [Breeze-ASR-25-BF16.gguf](https://huggingface.co/handy-computer/Breeze-ASR-25-gguf/resolve/main/Breeze-ASR-25-BF16.gguf) | 3.10 GB | 2.29% | 8.12% |

Discussions (0)

none scraped.

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