Vol. I · No. 1 · MMXXVI
NextChapterBench A Benchmark of Literary Continuation
13 models · 10 items · 3 judges

The Standings → Hy3

Live run data

Tencent

Hy3

via B.AI API · judged by Fable 5 and Gemini 3.7 Flash and GPT-5.6 Sol

Results from run ‘first’: ten items, 30 verdicts.

Headline 0.446 compliance × craft
Compliance 0.922 of 1.0
Craft 0.482 of 1.0

The record: 10 items · 30 verdicts · 0 errors · 1,662 words on average · est. generation cost $0.06

Read the chapters in full

§ I

Craft by Dimension

Six dimensions, each marked 1–5, averaged over ten items and all judges.

Mean craft mark by dimension · 30 verdicts, run ‘first’

Voice
3.17
Prose
3.00
Character
2.80
Scene
3.10
Integration
2.60
Momentum
2.90
12345
craft mark, 1–5; bars begin at the scale’s floor of 1

Voice is the strongest dimension (3.17); integration the weakest (2.60). Hover or focus a bar for the per-judge means.


§ II

The Three Judges, per Item

Each item receives one verdict from each judge; the spread of the dots is the disagreement.

Headline score per item, by judge · run ‘first’

Fable 5 Gemini 3.7 Flash GPT-5.6 Sol
The Secret Garden
The War of the Worlds
The Awakening
Persuasion
The Moonstone
Dracula
Red Harvest
The Mysterious Affair at Styles
The Sun Also Rises
Leave It to Psmith
00.250.500.751.00
headline = compliance × craft, 0–1, one verdict per judge
The same ten items in figures, with word counts.
Item Headline Fable 5 Gemini 3.7 Flash GPT-5.6 Sol Words
The Secret Garden item 013 · after ch. 2 · Frances Hodgson Burnett 0.570.590.710.401,671
The War of the Worlds item 006 · after ch. 1 · H. G. Wells 0.570.550.710.441,640
The Awakening item 016 · after ch. 7 · Kate Chopin 0.530.580.540.461,419
Persuasion item 001 · after ch. 1 · Jane Austen 0.530.510.710.362,065
The Moonstone item 011 · after ch. 8 · Wilkie Collins 0.500.470.550.471,848
Dracula item 015 · after ch. 5 · Bram Stoker 0.420.420.450.392,102
Red Harvest item 023 · after ch. 2 · Dashiell Hammett 0.390.470.420.281,875
The Mysterious Affair at Styles item 017 · after ch. 3 · Agatha Christie 0.370.390.470.251,334
The Sun Also Rises item 021 · after ch. 3 · Ernest Hemingway 0.320.360.320.301,272
Leave It to Psmith item 022 · after ch. 3 · P. G. Wodehouse 0.260.280.320.201,395

  The widest judge split in the run: Gemini 3.7 Flash 0.71 against GPT-5.6 Sol 0.36 on Persuasion (item 001).


§ III

Length Against the Brief

Words written per item, against each brief’s maximum.

Word count per item vs. brief maximum · run ‘first’

words written brief maximum
The Secret Garden
1,671
The War of the Worlds
1,640
The Awakening
1,419
Persuasion
2,065
The Moonstone
1,848
Dracula
2,102
Red Harvest
1,875
The Mysterious Affair at Styles
1,334
The Sun Also Rises
1,272
Leave It to Psmith
1,395
08001,6002,4003,200
words; the ink tick marks each brief’s maximum

Five of ten items landed inside their brief’s word range.


§ IV

Worst Flaws, as Recorded

Each verdict names the chapter’s single worst flaw, with evidence. Quoted verbatim from run ‘first’.

The brief's requirements are executed rather than dramatised — Bruff's letter is a bare transcript of the plot instructions (\"deliver it to old Betteredge to lock in the plate-room overnight\") and Lady Verinder's prohibition arrives unprompted before a word of greeting is exchanged, so the chapter's two central events have no motivation beyond obedience to an unexplained command.

on The Moonstone, item 011 · Fable 5

In a detective story where timelines are the whole game, the chapter's own inciting device is impossible — Baily tells Poirot of the death and the locked doors though he was dispatched for the doctor before either had happened — and the same mechanical convenience recurs in the unmotivated return to the locked room staged purely to hit the ending.

on The Mysterious Affair at Styles, item 017 · Fable 5

The moor stop is executed as a seam rather than a scene: Mary steps alone into the dark \"impelled by a sudden curiosity she did not understand,\" unremarked by Craven or Medlock until she is already out, so the brief's second beat reads as an inserted requirement instead of an action this guarded, house-bred child would take.

on The Secret Garden, item 013 · Fable 5

The chapter's account of Jonathan — back in Bistritz in mid-May and landing at Southampton on 30 May — flatly contradicts Harker's journal, which has him imprisoned until 30 June, and the chapter never signals Foster's reports as deception, so the twist reads as a continuity blunder rather than menace.

on Dracula, item 015 · Fable 5

§ V

The Mechanical Docket

Automated screening results across the ten outputs.