PulseAugur
中
实时 19:05:06
English(EN) Gradium Open-Sources Phonon, a 31M-Parameter On-Device TTS Model Under Apache 2.0

Gradium 开源 Phonon 设备端 TTS 模型

Gradium 发布了 Phonon,这是一个设备端文本转语音模型系列,提供 3100 万和 4000 万参数版本。仅限英语的 Phonon 31m 模型采用 Apache 2.0 许可,体积小巧,仅 64.5 MB,可以在包括 CPU、手机和浏览器在内的各种设备上本地生成语音,无需 GPU。Gradium 声称 Phonon 实现了具有竞争力的性能,其中 31m 模型在挑战性文本数据集上报告的词错误率为 1.70%,并在 CPU 上以毫秒级速度生成初始音频。 AI

影响 在各种设备上实现本地、低延迟的语音生成,减少对云服务的依赖。

排序理由 该集群描述了一家公司新文本转语音模型的开源。 [lever_c_demoted from research: ic=1 ai=1.0]

在 MarkTechPost 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

Gradium 开源 Phonon 设备端 TTS 模型

本文如何被排名

Signal score
1 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群描述了一家公司新文本转语音模型的开源。 [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
1 days old
Coverage has settled into its steady-state source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. MarkTechPost TIER_1 English(EN) · Asif Razzaq ·

    Gradium Open-Sources Phonon, a 31M-Parameter On-Device TTS Model Under Apache 2.0

    <p>Gradium has open-sourced Phonon, a family of on-device text-to-speech models. The English Phonon 31m ships under Apache 2.0, fits in 64.5 MB, and starts speaking in about 7 ms on a laptop CPU.</p> <p>The post <a href="https://www.marktechpost.com/2026/10/22/gradium-phonon-31m-…