AI & GEO
MUM (Multitask Unified Model)
Last updated
What is MUM (Multitask Unified Model)?
MUM (Multitask Unified Model) is an AI model Google announced in May 2021, which Google described as 1,000 times more powerful than BERT. MUM can understand and generate language across 75+ languages simultaneously, process text and images together (multimodal understanding), and handle complex, multi-step queries that would previously require multiple separate searches.
MUM enables Google to understand nuanced, cross-lingual, and multi-format queries — for example, a question asked with both a photo and text, or a multi-part planning question that requires synthesizing information across many sources. Google has said MUM is not used for general ranking; it is applied to specific tasks, such as recognizing variations of COVID-19 vaccine names across more than 50 languages and powering features like "Things to know".
Why it matters for SEO
MUM represents Google's capability to understand complex, multi-dimensional user needs that span topics, languages, and media formats. Content strategies that address comprehensive topical depth and anticipate the multiple sub-questions involved in complex user journeys are best positioned to align with how MUM interprets relevance.
Example
Google's announcement example was: "I've hiked Mt. Adams and now want to hike Mt. Fuji next fall, what should I do differently to prepare?". Answering it well requires comparing elevation, weather and gear, possibly using Japanese-language sources. A hiking site that covers how preparation differs between specific peaks addresses the kind of multi-part question MUM was designed for.
Go further with MUM (Multitask Unified Model)
Put the theory to work
LazySEO researches keywords, writes SEO articles and publishes them to your site on a schedule.