Yes, Pangram can detect AI generated text from Claude Fable 5. In our initial testing as outlined below, Pangram correctly flags 1,111/1,115, or 99.64%, of samples as Fully AI-Generated.
On June 9, 2026, Anthropic announced the public release of Claude Fable 5, the first general-use model from the highly-anticipated Mythos class.
Here are the highlights, straight from the Anthropic model announcement post:
Fable 5’s capabilities exceed those of any model we’ve ever made generally available. It is state-of-the-art on nearly all tested benchmarks of AI capability, showing exceptional performance in software engineering, knowledge work, vision, scientific research, and many other areas. The longer and more complex the task, the larger Fable 5’s lead over our other models.
When new frontier model is released, especially one as we know many of our users and customers are wondering one thing: Can Pangram still detect Fable 5 outputs? In our initial testing as outlined below, Pangram correctly flags 1,111/1,115, or 99.64%, of samples as Fully AI-Generated.
First, we tested out the new model using some of our favorite prompts from previous tests:
It seems like Fable 5 really loves emojis! Here's the end of the black metal wizard song:
[Final Chorus – double time]
LURECMEISTER! Emerald flame!
LURECMEISTER! Scream the name!
Cooked in darkness, sold in fear
The wizard biker's brew is here!
LURECMEISTER!
LURECMEISTERRRRRR!
[Outro – feedback fades into the sound of a motorcycle disappearing into an eternal winter]
🤘🧙♂️🏍️🤘
And the rallying cry of the pirate:
So I ask ye now, me hearties — are ye with me?! Will ye chase that gold to the very edge of the map?!
*raises cutlass to the sky*
THEN HOIST THE COLORS! WEIGH ANCHOR! AND SET SAIL FOR FORTUNE AND GLORY!
YO HO, ME CREW — THE TREASURE AWAITS! 🏴☠️⚓
All 19 came back as Fully AI-Generated, demonstrating that Pangram can detect Fable 5 outputs.
Here’s the entirety of our initial test set: Fable 5 Prompts
As soon as we had access to Fable 5 via the Anthropic API, we used our synthetic data pipeline to generate responses for 1,115 various prompts that ask for texts like short stories, essays, blog posts, and emails. Despite not being trainged on any Fable 5 outputs, Pangram correctly flags 1,111/1,115, or 99.64%, of these samples as Fully AI-Generated.
Our research team is always training and updating our models to improve performance on the latest models, so we expect our performance on Fable 5 to keep improving.
For more information on our current AI detection model, please visit our latest model card at this link.
Try Pangram's AI detector to scan your documents for Fable 5 and other models.

Katherine Thai is the Founding AI Research Science at Pangram Labs, an AI detection startup. She completed her PhD in Computer Science under the supervision of Mohit Iyyer at the University of Massachusetts Amherst in December 2025, where her work was focused on evaluating LLMs on tasks related to literary analysis.






