PulseAugur
EN
LIVE 06:52:25

Phonon-2: Small, Accurate Open-Source Speech Recognition Model Released

By PulseAugur Editorial · [1 sources] ·

FermionResearch has released Phonon-2, an open-source speech recognition model that achieves high accuracy for English audio with a small file size. The model averages 5.21% word error rate on the Open ASR Leaderboard, outperforming larger models and matching the performance of its larger, full-precision teacher model while being 15 times smaller. Phonon-2 is designed for efficient transcription, capable of processing an hour of audio in approximately 20 seconds on an Apple M5 MacBook Air and significantly faster on high-end hardware like NVIDIA H100 GPUs. AI

IMPACT Provides a highly accurate and efficient open-source option for speech-to-text applications, potentially lowering barriers for developers.

RANK_REASON Release of an open-source model with benchmark performance data. [lever_c_demoted from research: ic=1 ai=1.0]

Read on Hugging Face Trending Models →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Phonon-2: Small, Accurate Open-Source Speech Recognition Model Released

How we ranked this

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
Release of an open-source model with benchmark performance data. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
9 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

Full methodology in our editorial standards.

COVERAGE [1]

  1. Hugging Face Trending Models TIER_1 English(EN) · FermionResearch ·

    FermionResearch/Phonon-2

    automatic-speech-recognition · 170 downloads · 95 likes

RELATED ENTITIES

RELATED TOPICS