/Patent Tracker
GooglePatent applicationAI & softwareInterest 6/10

Google patents method for summarizing long audiobooks using AI

Google has filed a patent for a system that uses a large language model to create summaries of long audio content, such as audiobooks. Because the content is too long for the AI to process at once, the system splits it into smaller sections, summarizes the first part, and then uses that summary along with the next section to create a rolling summary of the entire file. The final summary can be shown as text or played as audio when requested by the user.

Why it mattersThis approach allows users to quickly grasp the main points of lengthy audio recordings without listening to the entire duration, which is useful for people who consume a lot of audio content.

In the filing’s own words

Implementations utilize a LLM to generate one or more summaries for long-form content such as an audiobook. Implementations determine whether the long-form content has a token length that exceeds a maximum token length for the LLM. In response to determining that the token length exceeds the maximum token length, the long-form content (or a transcript thereof) is segmented into at least a first content portion and a second content portion. A first summary is generated for the first content portion based on processing the first content portion as input using the LLM. A second summary is generated for the second content portion based on processing the first summary and the second content portion using the LLM. An overall summary can be generated to include the first and second summaries. The overall summary can be rendered visually and/or audibly in response to a user request for a summary of the long-form content.

What exactly they claim (claim 1)

1. A method implemented using one or more processors, the method comprising: determining that a token length of an instance of long-form content satisfies a threshold token length; segmenting, in response to determining that the token length of the instance of long-form content satisfies the threshold token length, the instance of long-form content into at least a first long-form content portion having a first token length and a second long-form content portion having a second token length, wherein the first token length and the second token length fail to satisfy the threshold token length; generating a first summary, summarizing the first long-form content portion, wherein generating the first summary comprises: generating a first text prompt that includes the first long-form content portion, that includes a first natural language instruction to generate a summary for the first long-form content portion, and that excludes the second long-form content portion; and causing the first text prompt to be processed using the large language model to generate the first summary for the first transcript portion; and generating a second summary, summarizing the second long-form content portion, wherein generating the second summary comprises: generating a second text prompt that includes the generated first summary that is summarizing the first long-form content portion, that includes the second long-form content portion, that includes a second natural language instruction to generate a summary for the second long-form content portion, and that excludes the first long-form content portion; and causing the second text prompt to be processed using the large language model to generate the second summary for the second long-form content portion; generating an overall summary for the instance of long-form content based on the first summary and the second summary; and causing rendering of the overall summary.

Facts

Company
Google (GOOGLE LLC)
Official title
UTILIZING GENERATIVE MODEL IN GENERATING SUMMARY OF LONG-FORM CONTENT
Publication number
US20260290331A1
Application number
19685962
Filed
May 22, 2026
Published
Sep 24, 2026
Inventors
2
Claims
20

Classified under

G10L 15/183 Physics & computingG10L 13/033 Physics & computingG10L 15/02 Physics & computingG10L 15/04 Physics & computing
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Google in AI & software

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.