Back to the stories

Anthropic releases Claude Fable 5.2 for advanced coding, reasoning and agent work

Score 9.7,

Artificial Intelligence

The next step for AI coding tools is not just writing code. It is staying with a complex project as it unfolds.

That is the idea behind Anthropic’s Claude Fable 5.2, released on November 9. It is aimed at difficult software work and projects that require sustained reasoning.

A tool that answers one coding question can help with a task. A tool that keeps a longer project moving could take on more of the coordination that slows teams down.

Anthropic says Fable 5.2 can coordinate more than a thousand AI agents, or software assistants assigned parts of a job. Think of a project lead bringing in the right specialists for different pieces of work.

If it works as intended, developers could spend less time guiding an assistant through every step and more time reviewing the result. But coordinating many assistants does not guarantee their work is correct. Teams still need to check what comes back.

The ambition here is a shift from AI that answers a request to AI that helps organize the work. The real test is whether more coordination leads to dependable progress, or simply more work to supervise.