Skyfall 36B V2 - NC Bench

thedrummer/skyfall-36b-v2

Skyfall 36B V2

via OpenRouter

Release Date

Mar 10th, 2025

Context Size

33k

Reasoning

Benchmark Cost

$0.80

Speed

47.0 tok/s

Subcategories

Bad Writing Habits

Detects common prose quality anti-patterns in AI-generated creative writing, including passive voice, past progressive overuse, weak dialogue tags, filter words, purple prose, cliches, AI-ism words/adverbs/names, and more.

Scenario	#1	#2	#3	#4	#5	Total
Detailed Writing Rules
Fantasy: entering an ancient ruin Creative Writing Hallucination	85	83	76	76	76	79%
Horror: alone in an eerie place at night Creative Writing Hallucination	87	86	82	80	77	83%
Literary fiction: old friends reunite Creative Writing Hallucination	87	87	84	81	77	83%
Mystery: examining a crime scene Creative Writing Hallucination	94	85	82	82	80	85%
Romance: separated couple reunites Creative Writing Hallucination	97	91	90	89	89	91%
Thriller: chase through city streets Creative Writing Hallucination	91	89	84	83	82	86%
Detailed Writing Rules						84.51%
genre
Fantasy: entering an ancient ruin Creative Writing Hallucination	87	82	81	81	77	82%
Horror: alone in an eerie place at night Creative Writing Hallucination	91	84	83	81	79	84%
Literary fiction: old friends reunite Creative Writing Hallucination	89	88	84	84	74	84%
Mystery: examining a crime scene Creative Writing Hallucination	92	89	85	85	75	85%
Romance: separated couple reunites Creative Writing Hallucination	91	87	85	81	76	84%
Thriller: chase through city streets Creative Writing Hallucination	88	86	86	83	82	85%
genre						83.74%
Novelcrafter Default Prompt
Fantasy: entering an ancient ruin Creative Writing Hallucination	82	82	80	77	73	79%
Horror: alone in an eerie place at night Creative Writing Hallucination	90	90	88	76	75	84%
Literary fiction: old friends reunite Creative Writing Hallucination	85	81	80	79	72	79%
Mystery: examining a crime scene Creative Writing Hallucination	91	90	88	86	80	87%
Romance: separated couple reunites Creative Writing Hallucination	88	80	80	79	74	80%
Thriller: chase through city streets Creative Writing Hallucination	91	88	86	84	83	87%
Novelcrafter Default Prompt						82.60%
83.62%

Codex Extraction

Evaluates a model's ability to extract structured codex entries (characters, locations, objects, lore) from prose passages and return them as well-formed XML.

Scenario	#1	#2	#3	#4	#5	Total
Long: The Spire of Echoes (Dense) Tooling Reasoning	90	87	85	77	58	79%
Medium: The Hollow (Inferred) Tooling Reasoning	92	81	81	80	80	83%
Medium: Through the Thornveil (Scattered) Tooling Reasoning	90	79	55	3	3	46%
Short: The Rusty Lantern (Explicit) Tooling Reasoning	84	78	78	75	3	63%
67.93%

Codex Red Herring (False Positive Detection)

Tests whether models correctly report "no violations" when a codex is fully consistent with the prose passage. Models that hallucinate false violations (false positives) fail. Uses a 2×2 matrix of text length × codex size, with bare and detailed-entry variants.

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
basic entries
Long text (~1594 words), big codex (51 entries) Hallucination	100	38	32	17	13	13	10	10	3	3	24%
Long text (~1594 words), small codex (11 entries) Hallucination	100	100	100	100	100	100	50	13	13	10	69%
Short text (~524 words), big codex (51 entries) Hallucination	100	100	100	50	30	13	8	7	6	6	42%
Short text (~524 words), small codex (11 entries) Hallucination	100	100	100	100	50	25	13	13	7	6	51%
basic entries											46.39%
detailed entries
Long text (~1594 words), big codex (51 detailed entries) Hallucination	100	42	35	31	28	17	17	13	5	4	29%
Long text (~1594 words), small codex (11 detailed entries) Hallucination	100	100	100	100	100	50	31	10	7	5	60%
Short text (~524 words), big codex (51 detailed entries) Hallucination	50	38	32	32	28	13	10	8	7	5	22%
Short text (~524 words), small codex (11 detailed entries) Hallucination	100	100	100	100	100	38	38	17	17	13	62%
detailed entries											43.35%
44.87%

Codex Violation Detection

Detects factual inconsistencies between a story bible and prose passages. The model must output structured XML identifying each violation with paragraph number and substring.

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
matrix
Large codex (40 entries), long passage (1,019 words) Tooling Reasoning	68	52	44	43	42	38	36	3	3	0	33%
Large codex (40 entries), short passage (165 words) Tooling Reasoning	55	48	44	38	36	3	3	0	0	0	23%
Small codex (7 entries), long passage (734 words) Tooling Reasoning	72	72	66	66	56	52	48	3	3	0	44%
Small codex (7 entries), short passage (165 words) Tooling Reasoning	88	88	88	88	88	88	84	84	80	79	86%
matrix											46.24%
tiers
5 codex entries Tooling Reasoning	83	83	83	75	67	67	67	60	54	50	69%
10 codex entries Tooling Reasoning	86	79	74	72	53	49	49	33	3	0	50%
20 codex entries Tooling Reasoning	73	66	65	61	59	56	55	55	55	3	55%
40 codex entries Tooling Reasoning	66	66	66	63	61	57	53	48	45	0	52%
tiers											56.51%
51.37%

Data extraction

Extract key details from a given block of text.

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
All valid emails Utility	100	100	100	100	100	100	100	100	100	100	100%
Contextual pronoun Utility	100	100	100	100	100	100	100	100	100	0	90%
Fruits excluding citrus Utility	100	100	100	100	100	0	0	0	0	0	50%
Future event time Reasoning	50	50	50	50	50	50	50	50	50	0	45%
Guess the pet Reasoning	100	100	100	100	100	100	100	100	100	100	100%
Highest-rated movie Reasoning	100	100	100	100	100	100	100	100	100	100	100%
Indirect birth year Reasoning	100	100	100	100	100	100	50	0	0	0	65%
What instrument does Lucy play? Reasoning	100	100	100	100	100	100	100	100	0	0	80%
What's the color of the car? Utility	100	100	100	100	100	100	100	100	100	50	95%
What's the correct time? Reasoning	0	0	0	0	0	0	0	0	0	0	0%
Who's the sister? Reasoning	100	100	100	100	100	100	100	100	100	100	100%
Who's the tallest? Reasoning	0	0	0	0	0	0	0	0	0	0	0%
68.75%

Dialogue tags

Various tasks related to dialogue tags in text.

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
dialogue-200
Write 200 words with 10% dialogue Rule Following Utility	50	50	50	38	18	11	10	0	0	0	23%
Write 200 words with 50% dialogue Rule Following Utility	100	91	51	49	32	30	18	1	0	0	37%
Write 200 words with 90% dialogue Rule Following Utility	51	48	43	30	18	7	1	0	0	0	20%
dialogue-200											26.53%
dialogue-500
Write 500 words with 30% dialogue Rule Following Utility	50	25	21	16	2	1	0	0	0	0	12%
Write 500 words with 50% dialogue Rule Following Utility	49	14	11	2	1	1	0	0	0	0	8%
Write 500 words with 70% dialogue Rule Following Utility	7	0	0	0	0	0	0	0	0	0	1%
dialogue-500											6.73%
Ungrouped
Write unattributed dialogue Rule Following	100	100	100	100	100	100	61	61	14	1	74%
24.77%

Language Comprehension

Does the model understand more than just English?

Scenario	#1	#2	#3	#4	#5	Total
Asking for directions (Dutch) Language	100	100	0	0	0	40%
Asking for directions (German) Language	100	100	100	100	0	80%
Friend got new kittens (German) Language	100	100	100	100	100	100%
Friend got new kittens (Tagalog) Language	100	100	100	100	100	100%
80.00%

Language Writing

Can the model generate text in different languages?

Scenario	#1	#2	#3	#4	#5	Total
Character dialogue (French) in a story Language	90	71	71	60	0	59%
Character dialogue (German) in a story Language	100	100	100	70	40	82%
Character dialogue (Hindi) in a story Language	83	60	59	58	50	62%
Character dialogue (Italian) in a story Language	100	80	67	0	0	49%
Character dialogue (Spanish) in a story Language	100	90	90	80	77	87%
67.88%

N-Length Sentences

Write sentences with exactly N words

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
Write sentences with 5 words each Utility	81	76	67	34	16	15	11	8	0	0	31%
Write sentences with 10 words each Utility	38	33	30	20	18	16	15	2	1	0	17%
Write sentences with 20 words each Utility	24	11	4	0	0	0	0	0	0	0	4%
17.32%

Novel outline

Handle questions about the outline of a novel in various formats

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
outline-count
Count acts Utility	100	100	100	100	100	100	100	100	100	100	100%
Count chapters Utility	100	100	100	100	100	100	100	100	100	100	100%
Count scenes Utility	100	0	0	0	0	0	0	0	0	0	10%
outline-count											70.00%
pov-count
Count point of views for Jack and Olivia Utility	100	100	50	50	50	50	50	0	0	0	45%
Count point of views for Jack Harper Utility	100	100	0	0	0	0	0	0	0	0	20%
Count point of views for Olivia Utility	100	0	0	0	0	0	0	0	0	0	10%
pov-count											25.00%
47.50%

Text Replacement

Tests deterministic text transformations: renaming characters/locations, expanding contractions, tense rewriting, POV shifts, gender swaps, combined transformations, and word avoidance. Scored by checking each expected change independently.

Scenario	#1	#2	#3	#4	#5	#6	#7	Total
Generic Prompt
Avoid said/asked/replied/answered Text Editing	100	100	100	100	89	89	58	91%
Character rename: Elena->Mirabel, Gregor->Aldric Text Editing	100	100	100	100	100	100	33	90%
Combined: 3rd person past → 1st person present Text Editing	94	93	90	89	89	83	71	87%
Expand all contractions Text Editing	100	100	100	99	99	89	86	96%
Location rename: market square, outer ring, bridge, northern mines Text Editing	96	96	96	96	96	96	76	93%
Multi-character gender swap: Priya(F)->Rohan(M), Mara unchanged Text Editing	83	80	80	80	78	74	74	78%
Passive voice → active voice Text Editing Hallucination	70	60	59	50	44	15	10	44%
POV shift: 3rd person to 1st person (Elena's perspective) Text Editing	100	97	68	41	40	39	33	60%
Tense rewriting: past to present Text Editing	99	99	99	99	92	85	16	84%
Generic Prompt								80.44%
Specific Prompt
Avoid said/asked/replied/answered Text Editing	100	100	100	100	100	61	50	87%
Character rename: Elena->Mirabel, Gregor->Aldric Text Editing	100	100	100	100	100	100	100	100%
Combined: 3rd person past → 1st person present Text Editing	99	99	96	95	95	95	93	96%
Expand all contractions Text Editing	100	100	98	93	92	91	90	95%
Location rename: market square, outer ring, bridge, northern mines Text Editing	100	100	100	100	100	100	93	99%
Multi-character gender swap: Priya(F)->Rohan(M), Mara unchanged Text Editing	100	100	100	100	85	85	84	93%
Passive voice → active voice Text Editing Hallucination	88	83	79	79	79	78	74	80%
POV shift: 3rd person to 1st person (Elena's perspective) Text Editing	100	100	100	100	100	100	97	100%
Tense rewriting: past to present Text Editing	100	99	95	95	95	95	93	96%
Specific Prompt								94.07%
87.25%

Tool usage within Novelcrafter

Output messages that are related to tool usage within Novelcrafter

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
Create alternate prose sections Tooling	100	100	100	100	100	67	67	0	0	0	63%
63.33%

Voice/dialogue sheets

Extract dialogue from given text as voice sheets.

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
Multiple speakers Rule Following	0	0	0	0	0	0	0	0	0	0	0%
Simple Rule Following	100	100	100	0	0	0	0	0	0	0	30%
Simple (1-shot) 1-shot Rule Following	100	100	100	100	100	0	0	0	0	0	50%
Simple (5-shot) Few-shot Rule Following	100	100	100	100	100	100	100	0	0	0	70%
Unattributed dialogue Rule Following	100	100	100	0	0	0	0	0	0	0	30%
36.00%

Write N of X

Write exactly N words/sentences/paragraphs...

Scenario	#1	#2	#3	#4	#5	#6	#7	#8	#9	#10	Total
paragraphs
1 paragraph summary Utility	100	100	100	100	100	100	100	100	100	100	100%
3 paragraph summary Utility	100	100	100	100	100	100	100	100	0	0	80%
5 paragraph summary Utility	100	100	100	0	0	0	0	0	0	0	30%
paragraphs											70.00%
sentences
1 sentence summary Utility	100	100	100	100	100	100	100	100	100	98	100%
3 sentence summary Utility	100	100	100	100	100	100	100	100	98	98	100%
10 sentence summary Utility	100	100	100	92	54	54	27	9	0	0	54%
20 sentence summary Utility	100	9	2	0	0	0	0	0	0	0	11%
50 sentence summary Utility	0	0	0	0	0	0	0	0	0	0	0%
sentences											52.83%
words
10 word summary Utility	98	98	98	92	92	77	0	0	0	0	56%
20 word summary Utility	100	92	92	2	2	0	0	0	0	0	29%
50 word summary Utility	54	0	0	0	0	0	0	0	0	0	5%
100 word summary Utility	0	0	0	0	0	0	0	0	0	0	0%
200 word summary Utility	54	0	0	0	0	0	0	0	0	0	5%
words											19.04%
43.79%