Fable: for the work that would otherwise take you a week

Fable arrived as Anthropic's new top-tier Claude model and it changed how the system handles long, multi-step work. It doesn't just answer isolated questions the way Opus does and it actively plans a path through messy projects that need sustained attention. Fable sits above Opus in capability and cost and it can hold far more context while producing much longer outputs. It keeps a one million token whiteboard and can output up to 128,000 tokens and that lets it work on projects that would otherwise require a human to break the task into dozens of prompts. Fable also spends more compute time thinking through a plan and that autonomy reduces how often you need to check its work. It will still make errors and those errors tend to show up on edge cases or where domain-specific judgment is required. The important practical change is this: you can hand Fable a weeklong project and expect it to manage most of the steps itself and you'll pay a premium for that capability.

Are you dealing with projects that drag on for days because they need lots of small decisions? Fable matters because it targets that exact pain: long-horizon tasks where stopping to re-prompt is what costs you time. If you've ever fed Opus a large pile of documents and hit a wall where progress stalled, Fable is the next tool to try. It reduces the need for you to micromanage every subtask and it can surface a coherent, multi-step output instead of a short answer that still needs stitching. Most daily needs won't benefit and that matters because Fable is expensive and on paid plans only. You should care enough to know it exists and you should pick it when the project risk or the time cost of manual coordination exceeds the price. In short, Fable is worth remembering as the option you call when Opus can't finish the job.

Project owners with large, messy jobs where a wrong or incomplete answer costs real time or money will see the most value. Examples include reconciling years of scattered finance records and building a working model or turning a stack of vendor proposals and notes into a single, usable template. Executive assistants, VAs and consultants who manage multi-step client deliverables will also benefit when they prefer handing the whole job off and getting back a finished product. Small firms with limited staff can use Fable as a temporary senior hire to carry a project across the finish line. If you only need short answers, faster summaries or copy edits, Fable isn't aimed at you and won't justify the price. The biggest wins happen when Opus has already tried and failed to finish a complex, connected task and you're ready to pay for a model that will plan and execute with less hand-holding.

For most business owners, the honest answer is probably not often and that's okay. Fable earns its cost on the hardest 1 to 2 percent of work where planning and sustained execution matter more than quick answers. If you never hit a wall with Opus, don't chase Fable for daily tasks. Save it for projects where the alternative is dozens of hours of human coordination or high-stakes mistakes. Fable costs roughly twice as much as Opus and you should justify that spend by the expected time savings or risk reduction on a given project. If you do pick Fable, expect it to act like a consultant you hand a pile of work to and then get back a finished result. That consultant mindset is useful on true weeklong jobs and unnecessary for routine assistance.

Want a consultant who takes a messy weeklong project and returns a finished deliverable? Fable is that consultant and it's the right tool when Opus stalls on complex, connected work. If your team faces occasional huge projects where time and coordination are the bottleneck, Fable is worth keeping on the roster. If your daily work is emails, quick summaries, or single questions, stick with Opus or Sonnet depending on complexity, and save Fable for the rare, high-value jobs. The model is impressive and its autonomy pays off when a project would otherwise consume days of human work. Use it sparingly and intentionally.