• "Le Chaton Fat," a fictitious Mistral AI model, has become the latest obsession in tech circles.
• The viral meme emerged after Mistral rebranded its Le Chat chatbot as Vibe, sparking backlash.
• AI leaders, including Replit CEO Amjad Masad, joined the cat craze.

Tech insiders can't stop talking about "Le Chaton Fat," a supposedly world-beating AI model from Mistral. There's just one problem: it seems to be fake.

The meme exploded over the past few days after a series of social media posts claimed that French AI startup Mistral had secretly developed a frontier model called "Le Gros Chaton" — French for "the fat kitten" — that outperformed rivals from OpenAI and Anthropic.

The joke has spread so widely that some people outside the AI bubble have started asking whether the model is real.

According to French technology news outlet Numerama, Le Chaton Fat — which mixes English and French spellings — doesn't exist; it's an elaborate running gag that appears to have emerged from Mistral's online community before spreading across AI circles.

Mistral AI didn't respond to Business Insider's requests for comment.

How the meme took off

While the meme's exact origin is unclear, jokes started emerging shortly after Mistral's CEO, Arthur Mensch, announced at its first-ever summit in Paris last month that it was rebranding its chatbot, Le Chat, as Vibe.

The move sparked debate among users, with some lamenting the loss of what they viewed as one of AI's most distinctive product names in posts on Reddit.

As the discussion snowballed, Reddit users began inventing elaborate feline-themed successors to Le Chat. Posts on the Mistral subreddit advertised the arrival of "Le Chaton Fat," a fictional frontier model described as having "30T+ params," a measure of a model's accuracy, "1000 meows per second," and "maximum chonk."

The joke quickly spiraled.

Social media users created fake launch announcements, fabricated benchmark charts showing Le Chaton Fat crushing rival models from OpenAI and Anthropic, and mock regulatory notices saying the European Union had barred the model because it was simply "too heavy to regulate."

The French Goverment needs to STOP Le Chaton Fat before it's too late. This is the FAT takeoff we've been warned against for years! pic.twitter.com/VoaBPJmfBT

— GLIF (@heyglif) June 15, 2026

A tongue-in-cheek poke at Anthropic

The spread of the meme appeared to be fueled by the US government's decision to bar any foreign entities or individuals from using Anthropic's cybersecurity-focused models, Mythos 5 and Fable 5.

The restrictions became a recurring theme for meme creators, who increasingly portrayed Le Chaton Fat as a powerful European alternative to Anthropic's newly restricted models.

pic.twitter.com/CF4OQIkaZV

— Alejandro (@vskblue) June 16, 2026

Ethan Mollick, professor and co-director of the Generative AI Lab at the Wharton School, appeared to reference it in an X post on Monday.

"The le chaton fat meme is leaking to the outside world and I expect to be asked about Mistral's new ginormous cat model with infinite benchmark scores at my next meeting with corporate leaders," Mollick wrote.

He later added: "Also better than being asked about wet Claude," referencing a viral meme that portrayed Anthropic's chatbot as unusually emotional and introspective.

Amjad Masad, the CEO of coding firm Replit, also joined the joke, posting on X on Tuesday: "Who needs Fable when you can have Mistral's Le Chaton Fat."

Even Mistral AI's CEO got in on the trend, saying in an X post on Monday: "It's actually le gros chaton"

A timely moment for Mistral

Some of the viral memes highlighting the US restrictions on Anthropic's new models could play into Mistral's long-standing argument that Europe should not become dependent on American AI providers.

Mistral has positioned itself as an alternative, championing open-weight models that customers can deploy on their own infrastructure and customize using their own data.

The US' export controls reinforced a warning Mensch has repeatedly made: reliance on foreign AI providers means access to a model can be restricted overnight.