Two new text-to-speech models, Gemini 3.8 Flash TTS and Flash-Lite TTS, turn voice generation into a creative studio for developers and creators. Flash TTS enables building bespoke character voices from natural language prompts, with granular, line-by-line direction of pacing, emotion, dialect shifts, backchanneling and nonverbal cues; it supports native two-speaker scene staging and long-form generation with minimal speaker drift. Flash-Lite is optimized for high-volume, cost-efficient dubbing, audio content creation and scalable voice agents. Both models run across Google AI Studio, Gemini API/Enterprise/Notebook and Google Vids, and expand an available library from 30 starter voices to 2,000+ production-ready voices across 100+ languages and dialects. Voice remixing (fine-tuning timbre, pitch, pace and accent) is noted as coming soon.
Performance claims include top rankings on Hume AI’s Voice Design Benchmark (#1 overall, 71.4; #1 in accent modeling, 60.8) and #1/#2 on Hume’s Overall Quality Index, plus leading blind-preference results in Voice Arena for multiple languages (Japanese, Brazilian Portuguese, Vietnamese, MSA, Mexican Spanish, Hindi). Practical features include voice replication from a 30-second sample with built-in consent verification, SynthID watermarking and C2PA credentials for provenance, plus tools for saving, managing and scaling custom voices for audiobooks, games, podcasts, dubbing and interactive voice agents.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Vibe Coding Production Kit is a workflow designed for developing AI coding agents, streamlining the process from issue identification to code deployment. It integrates with tools like GitHub Copilot and offers resources for developers working on AI-powered software development. (github.com)
SkillOpt treats natural-language skills as trainable external artifacts for frozen agents, optimizing them through rollouts, reflection, and validation. It improves agent performance by making the skill document itself the target of structured edits, avoiding fine-tuning of the underlying model. (microsoft.github.io)
Opus 5.5 can generate explainer videos automatically using a serverless agent on OpenComputer, taking a URL or prompt to produce a video in about four minutes. It leverages a microVM environment, web tools, and a deterministic rendering process to create high-quality videos without manual editing. (launchvideo.io)
Anthropic has signed a $12 billion agreement with Akamai to access AI computing resources. The deal aims to support Anthropic's development of advanced artificial intelligence models. (bloomberg.com)
Tech leaders from major AI companies have urged the United Nations Security Council to regulate artificial intelligence technology to prevent potential threats to humanity. They warn that unchecked AI development could pose significant risks to global safety. (apnews.com)
GPT-6 Sol shows significant performance degradation compared to previous models, with lower test scores and many hidden bugs. GPT-6 Luna performs even worse, indicating a decline in quality for the new release. (twitter.com)
Today's best Hacker News stories, summarized and screenshotted, one email a day.