KM4BI

Claude Fable 5.1: More Than Code — a Film, 7 Sheets and 13 Slides

A test of Fable 5.1 across modeling, writing and Blender shows the trade-off between speed, diligence and verifiability in knowledge work.

Published

Fable 5.1 is tested on work whose quality is harder to validate than code that either runs or fails: a financial workbook, an executive deck, concise writing and a visual sequence. The central lesson is procedural: create a strong first version quickly, then raise the effort level and scrutiny when the stakes justify it.

From a useful draft to deeper diligence

For a GoPro–Starman post-acquisition exercise, the “low” setting produced a seven-sheet workbook and a 13-slide deck. The formulas and scenarios worked, but sources and checks were not easy to audit. At “extra,” the model expanded its diligence: nine sheets, 15 slides, 26 linked sources, explicit funding needs and close probability, plus an exit-multiple check.

The generated valuations are not presented as ground truth; the available information was incomplete. The more important observation is that a finished-looking deliverable still needs visible assumptions and verification.

Complementary models rather than one winner

In the comparison, GPT-5.6 Soul stood out for inspectability, with dedicated sources and an explicit checks sheet. Fable 5.1 was especially strong at rapid production and visual presentation. In a constrained 100-word writing task, Fable 5.1 improved causal clarity relative to Fable 5, while Soul favored a smoother narrative.

That makes a sequential workflow plausible: fast first pass, review of assumptions and disagreements, structural checking, then polish. For knowledge work, the first draft is not the end of the process.

From an idea to a Blender film

Given only an address and a loose brief, Fable 5.1 built an architectural model in Blender and a 37-second film. It created terrain, interiors, lighting and a camera path, inspected renders, adjusted scenes and rendered the final sequence. The result has visual limitations, but it demonstrates how an idea can be made tangible without direct mastery of the 3D tool.

Why it matters

Lowering the effort needed to create a workbook, deck or visualization makes exploration cheaper. At the same time, it raises the importance of checking sources, formulas, assumptions and decision criteria. The effort dial becomes an operational choice: “low” to explore and “extra” to deepen work when the cost of being wrong is higher.

Source