Mistral releases Large 4, a new frontier reasoning model

Simon Willison published a Hacker News comment on October 6 responding to a thread titled “Mistral Large 4.” Willison’s post links to his own remark within that discussion and quotes a comment from user wren6991, who wrote, “The benchmark is saturated. Frontier models are tested with an armadillo in fishnet tights jaywalking on Mars.”

Willison’s post indicates he found the thread’s content irresistible enough to reply, and includes a code fragment referencing a command using “llm -m claude-opus-5.5,” suggesting some kind of command-line tool invocation involving a model named Claude Opus 5.5. The provided material does not elaborate on the content or outcome of that command, nor does it specify what Willison’s own comment said beyond the link itself.

The source material does not establish when Mistral Large 4 was released, nor does it characterize the model’s capabilities or classification beyond its appearance in the thread’s title. The material likewise does not clarify what benchmark wren6991 referred to, nor the context of the “armadillo in fishnet tights jaywalking on Mars” remark beyond the quoted text itself.

This item originates from a Hacker News discussion thread and a brief accompanying post by Simon Willison dated October 6, 2026.