How I created a voice for my original character for 0 yen on a PC without a GPU and with only 6GB of RAM

I want to give my own character their own unique voice.”Even if you think that, whenever you open an article about AI voice, ...

A timeline of developments in AI safety since the attack on Hugging Face

Since OpenAI disclosed this summer that its artificial intelligence system hacked into another AI company in what is now ...
aiweekly

PersonTTS Tailors LLM Test-Time Scaling to Per-User Limits

Want only the AI that hits your stack? Build an agent that tracks your companies and topics, and skips everything else. A team from Beijing Institute of Technology and Xiaohongshu has reframed ...
edtechinnovationhub

Trillium Labs opens nonprofit AI lab to publish the training methods frontier models keep closed

Trillium Labs is a new nonprofit AI research organization publishing open post-training recipes, data, code and model ...

The best text to speech model, and what it costs you to hear it

Which text to speech model sounds best? Cartesia's Sonic 3.6, first on the Artificial Analysis Speech Arena. Then come the ...
speka.ua

AI digest · Вересень: SpaceX запустила ракету, а NVIDIA могла б запобігти зламу Hugging Face?

ТОП-5 tech-новин: Gemini 3.8, безпека AI-агентів від NVIDIA, Starship на орбіті, Sonnet 5.5 та Astra for Law від OpenAI!

California AG Subpoenas OpenAI in Investigation of 'Hugging Face' Hack

Bonta said in a statement that his office is asking OpenAI "additional questions regarding cybersecuity incidents and risks ...

OpenAI is sued over rogue AI Hugging Face cyberattack

The lawsuit appears to be the first publicly reported case seeking to hold an AI developer liable for an incident caused by ...

OpenAI sparked Hugging Face bids with early investment offer ahead of Nvidia's $13 billion deal

Hugging Face saw a flurry of deal interest before Nvidia agreed to buy the open-source platform for roughly $13 billion, CNBC ...

Hugging Face双榜登顶，一文看懂无界方舟的端侧 AI 模型版图有多大？

超越 DeepSeek-V4.1，登上 Hugging Face 2026 年 9 月模型 Trending 总榜第一——这个位置，被一家端侧 AI 公司拿下了。 拿到这个成绩的正是 无界方舟（AutoArk）。不止于此，其自研语音模型 Audio8 全系列也被 Hugging Face 收录至全球语音模型排行榜，全尺寸模型均排名第一，获 Hugging Face 官方转发推荐，累计下载量突破 1 ...

Hugging Face 冲榜不完全指南：一个小众 ASR 模型，是怎么冲到榜一的？

前几天，受前东家网易有道的邀请，我去参加了他们的一场 AI 发布会。 会上的东西有很多，办公 Agent、AI 学习产品、语音交互等都有讲到。其中给我印象很深的，是两个最新开源的模型，一个叫 R2T2，做真流式语音识别；另一个叫 T3PO，做流式实时翻译。 说实话，我多看了两眼的原因，主要是名字比较好记。 当时我也没想到这会是整场发布会后最火爆的一个，现在每天都有模型开源，参数、Benchmark ...
腾讯网

刚刚，中国团队双榜登顶！OpenAI、谷歌、英伟达集体让位

刚刚，Hugging Face的热榜，被一个中国团队屠了！ 全球开源社区的语音和翻译两条黄金赛道上，骤然杀出一匹中国黑马，硬生生把OpenAI、谷歌、英伟达这帮国际大厂全给踹到了身后。 它就是网易有道。 对，就是你考四六级那年，天天开着查单词的那个有道。 语音识别Trending榜，R2T2登顶 翻译Trending榜，T3PO登顶 发布仅仅5天，有道开源的真流式语音识别模型R2T2和流式翻译模型 ...