14.4B tokens in four weeks: developer ships 65 projects testing Fable 5.2, ~$10,300 in API costs
kieranklaassen · x · 2026-11-10
- Developer hammermt spent 14.4B tokens ($10,300 at Fable 5.2 API prices) over four weeks testing the model and published a build log of 65 shipped projects.
- Token breakdown: a 3D "Battle of New York" from a single prompt consumed 4.7B (33%), HIPPO Board Canvas 2.1B (14%), What is Checks 1.8B (12%), with 53 other projects totaling 2.5B (17%).
- Measured from 42 Claude Code sessions on a Mac (7.9B) plus estimates where transcripts were missing (6.5B) — a rare project-level account of large-scale vibe coding spend.
More from coding & agent
- Community-built Claude skill prunes useless UI copy, works surprisingly well with Opus — floguo · 2026-11-10
- Two 'fresh eye review' prompts catch AI coding mistakes after generation — doodlestein · 2026-11-10
- InstructGPT co-author's Jev harness PDF promises 200x faster, 400x cheaper coding agents — Redo · 2026-11-10
- Three coding-agent security patches ship this week: Mistral Vibe, Jan, DeepSeek Harness — alvelda · 2026-11-10
- Dev rewrote Node app in Go, asks why developer-friendly languages matter when AI writes code — rakyll · 2026-11-10
- Claude Fable 5.2 Lands in Factory: Works Through Diagnostics, Keeps Precision — matanSF · 2026-11-10