DEV Community

Cover image for The Fable 5 Receipts: What AI Coding Shipped
Max Quimby
Max Quimby

Posted on • Originally published at agentconn.com

The Fable 5 Receipts: What AI Coding Shipped

The conversation about AI coding shifted this week. Not because of a benchmark. Not because of a demo. Because three practitioners showed their receipts.

📖 Read the full version with charts and embedded sources on AgentConn →

Simon Willison published the exact cost of shipping an open-source release mostly written by Claude Fable 5: $149.25. A developer ported a 23-year-old game engine to the iPhone in under 19 hours. And Stripe reportedly ran a 50-million-line Ruby migration with Fable 5 in a single day — work their team estimated would take two months.

These aren't toy demos or benchmark scores. They're production artifacts with costs, commit histories, and deployment logs.

Receipt #1: The $149.25 Invoice

On July 5, Simon Willison published a blog post with a title that reads like a line item: "sqlite-utils 4.0rc2, mostly written by Claude Fable (for about $149.25)."

sqlite-utils is a widely-used Python library for manipulating SQLite databases. The 4.0 release candidate is a significant refactor with migrations, nested transactions, and breaking changes. Willison published the exact token costs, documented each session, and noted where Fable excelled and where it needed correction. During a final review, Fable identified five significant problems categorized as release blockers — issues Willison hadn't caught in his own passes.

The $149.25 figure represents the total API spend. At $10/$50 per million tokens, that's roughly the cost of three hours of a mid-level developer's time.

Hacker News thread discussing Simon Willison's sqlite-utils 4.0rc2 release mostly written by Claude Fable for $149.25

Receipt #2: The 19-Hour Port

Developer Ammaar Reshi posted a video on X showing Command & Conquer: Generals Zero Hour running natively on an iPhone. Not through an emulator. The actual C++ engine, compiled for ARM64, with touch controls and pinch-to-zoom.

The project was possible because EA released the source code under GPL v3. The GeneralsX community had spent months on the macOS port. Reshi used Fable 5 via Claude Code for the iOS leap: adapting the engine, building touch input handlers, and wiring up MoltenVK.

The first build took 40 minutes. The entire project landed in 19 commits over 19 hours. Opus 4.8 couldn't complete the same task even at its highest effort setting.

Hacker News thread with 633 points discussing C&C Generals ported to macOS iPhone iPad using Fable

Receipt #3: The 50-Million-Line Migration

Anthropic published a customer signal from Stripe: Fable 5 performed a codebase-wide migration on their 50-million-line Ruby codebase in a day, where the manual path would have taken over two months.

⚠️ The "50 million lines in a day" framing deserves scrutiny. Stripe's team still had to validate every change against live environments. The implementation time compressed. The review time didn't. "The ratio of implementation time to review time inverts." That's important — but it's not "an AI wrote 50 million lines of code."

The Bottleneck Shift

A Fable 5 engineer articulated: "You, not the model, are now the limit."

Where AI coding works right now:

  • Legacy code transformations with clear structure
  • Pattern-based migrations across large codebases
  • Well-scoped library work with a knowledgeable operator
  • Reverse engineering with tool integration

Where it doesn't — yet:

  • Greenfield architecture decisions
  • Anything where the operator can't evaluate the output
  • Novel algorithm design
  • Systems where correctness matters more than coverage

What This Means for Coding Agent Stacks

At $10/$50 per million tokens, a day of heavy Fable 5 usage costs $200–$400. After July 7, Fable 5 moves to usage-based credits. Teams that built workflows around uncapped access will face cost discipline for the first time.

The receipts this week don't change the picture. They sharpen it. The model is genuinely better. The operator still matters more.

Originally published at AgentConn

Top comments (0)