Well that would depend on the definition and what you exactly mean by untyped.

The untyped part is usually referring to the way the programmer interacts with the language, for example not setting a type for variables and parameters. But then there is the question of is the programmer ever allowed to explicitly set the type. And further more, if the programmer explicitly set the type, does this mean the type can't change at a later point? And another question could be, can the programmer check or enforce what type a variable or parameter is? And the question, if there is only one type of data in the language, would that be a typed or untyped language? But I would consider these to be details and all fall under the untyped umbrella, with untyped just meaning not-typed.

Then there's the question of the technical implementation of the language. Defining a language is one thing, actually having it run on a real system is another. Usually technical systems at some point require explicit types. Something somewhere needs instructions on how to handle the data and this usually leads to some kind of typing instructions being added along with the data. But depending on how many abstraction layers there are, this can soon become a very pedantic discussion. I feel what matters is the design, definition and intend of a language. The actual technical implementation isn't what matters in my opinion.

I feel like there are so many programming languages and technical systems at this point, every variation and exception exists. And if you can think of one that doesn't exist, expect a follow up comment of somebody pointing out it does exist after all, or them having started a project to make it exist in the near future.