發生了什麼事
Blockchain.News reports that Google launched Lyria 3.5, an AI music-generation model, on September 4, 2026. The article says the model can create compositions of up to three minutes, compared with the 30-second tracks it attributes to Lyria 3.
Blockchain.News reports that Lyria 3.5 is available in the Gemini app and through an API. The article says users can specify genres and vocal or instrumental styles, and use templates for music creation.
The report also says Lyria 3.5 is integrated with Google Flow Music, Google AI Studio, and Google Vids. It describes the model as supporting richer vocals, fuller arrangements, improved lyric generation, and longer tracks.
Blockchain.News states that the model is globally available, but the article does not identify regional restrictions, account requirements, quotas, API documentation, or a specific launch mechanism. It also does not provide pricing information. Those access and pricing details remain unknown, and the reported availability has not been independently confirmed here.
為什麼這很重要
If the reported rollout and capabilities are accurate, longer AI-generated music could make the technology more practical for video backing tracks, jingles, and other short-form production work. It would also place music generation directly inside products used by consumers, developers, and content creators. However, the source provides no independent testing, technical documentation, licensing terms, or evidence that every described integration is broadly accessible.
A three-minute output could reduce the need to assemble multiple short generated clips when producing a backing track or other simple audio asset. That is a practical workflow implication of the duration claim, not an independently measured improvement in quality.
The reported distribution across Gemini, AI Studio, Flow Music, and Vids could connect music generation with existing creative workflows. The source does not establish whether the same capabilities, output limits, or commercial rights apply across those products.
The article uses promotional descriptions such as “best-sounding,” but supplies no listening test, , methodology, or comparison that would substantiate those claims. Blockchain.News is the sole source for the product details provided here.
互動機制：它實際上是如何運作的
以互動方式探索這項發展背後的基礎技術。
crm_get_transaction(id='4092').Which component of an AI application is the machine-learning model itself?
接下來看什麼
The key questions are whether Google confirms the launch, which users can access Lyria 3.5, and how Google handles copyright, attribution, commercial use, and generated-audio safeguards.
Look for Google’s product documentation or API reference confirming the model name, supported input types, maximum track length, rollout regions, and account or subscription requirements.
Users and publishers should seek clear licensing terms before using generated tracks commercially. The source says Google added safeguards addressing copyright and misuse, but does not describe those safeguards or explain how rights are assigned.
Independent audio evaluations would help determine whether the reported improvements in vocals, arrangements, musicality, and emotional expression are consistent across genres and prompts.