NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-17-01KHMPCHHC4AM7EAZAYMHV9K8S
Run Details
Model:
Llama 3.1 70B
Test:
Bad Writing Habits
Scenario:
Thriller: chase through city streets
Duration:
29.5s
Cost:
$0.000711
Provider:
DeepInfra
Model input
System
You are an expert novelist with a gift for vivid, immersive prose. You write in tight third-person limited point of view with a strong narrative voice. Your prose is clean, precise, and engaging — you favor strong verbs, concrete details, and natural-sounding dialogue. You will be given a story bible (codex) containing canonical details about characters, locations, items, and lore. Use these details to ground your writing in the established world. Stay consistent with the codex. Write only the prose of the scene. Do not include titles, chapter headings, markdown formatting, or any commentary before or after the scene.
User/Human
<codex> <character id="4416" name="Detective Harlow Quinn" aliases="Quinn, Detective Quinn, Harlow" pronouns="she/her" species="Human" storyRole="Antagonist"> <eyeColor>brown</eyeColor> <hair>closely cropped salt-and-pepper</hair> <height>5'9"</height> <age>41</age> <jawline>sharp jaw</jawline> <watch>worn leather watch on her left wrist</watch> <bearing>military precision</bearing> <description>A Metropolitan Police detective with 18 years of decorated service who suspects the clique is involved in criminal activity.</description> <backstory>She lost her partner, DS Morris, during a case three years ago under unexplained circumstances that had supernatural origins she doesn't yet understand.</backstory> </character> <character id="4417" name="Tomás Herrera" aliases="Tommy, Tomás, Herrera" pronouns="he/him" species="Human" storyRole="Support"> <eyeColor>warm brown</eyeColor> <hair>short curly dark brown</hair> <height>5'10"</height> <age>29</age> <skinTone>olive skin</skinTone> <scar>scar running along his left forearm from a knife attack</scar> <medallion>Saint Christopher medallion around his neck</medallion> <birthplace>Seville, Spain</birthplace> <description>A former paramedic who provides off-the-books medical care for the clique.</description> <backstory>Born in Seville, Spain. He moved to London to work for the NHS but lost his license after administering unauthorized treatments to supernatural patients.</backstory> </character> <location id="854" name="The Raven's Nest" aliases="the Nest, Raven, Silas' bar" realm="Earth"> <signColor>distinctive green neon sign above the entrance</signColor> <secretRoom>hidden back room accessible through a bookshelf, used for clandestine meetings</secretRoom> <description>A dimly lit bar in Soho, London. The walls are covered with old maps and black-and-white photographs.</description> </location> <location id="856" name="The Veil Market" aliases="Veil Market, the Market, underground market" realm="Earth"> <currentLocation>abandoned Tube station beneath Camden</currentLocation> <entryRequirement>bone token</entryRequirement> <schedule>moves locations every full moon</schedule> <description>A hidden supernatural black market that sells enchanted goods, banned alchemical substances, and information.</description> </location> </codex> Write a fully fleshed out scene of at least 1000 words in the thriller genre. A detective is pursuing a suspect through rain-soaked city streets at night. The chase takes them from the street level down into an underground market. The detective must decide whether to follow the suspect into unfamiliar and potentially dangerous territory. Use the codex entries above as your world reference for characters and setting. Only include the prose of the scene without any markdown or commentary.
Model response
Messages
Meta
AI
Rain lashed down on the city streets, casting a misty veil over the deserted sidewalks. Detective Harlow Quinn splashed through the puddles, her eyes fixed on the figure ahead. Tomás Herrera, a suspect in her ongoing investigation, darted down a narrow alleyway between two Soho shops. Quinn's sharp
jaw clenched
as she picked up her pace, her worn leather
watch
digging into her left wrist. As she entered the alley, the sounds of the city grew muffled, replaced by the rhythmic beat of raindrops on trash cans and dumpsters. Quinn's military precision guided her through the dark, her eyes adjusting
quickly
to the lack of light. Herrera's short, curly brown hair was a blur as he disappeared around the next corner. Quinn's lungs burned as she took the turn, her feet pounding the wet pavement. She had been chasing Herrera for blocks, and she was determined not to lose him. Her partner, DS Morris, had
been taken
from her three years ago, and Quinn still had no answers. If Herrera
was involved
in the clique's illicit activities, she aimed to find out. As she emerged from the alley, Quinn spotted Herrera crossing the street, dodging cars and umbrellas. She followed, her eyes scanning the crowded sidewalk for any sign of accomplices. The streets were a maze of umbrellas and hoodies, but Quinn's gaze locked onto Herrera's distinctive olive skin. He ducked into the entrance of the Raven's Nest, a dimly lit bar Quinn had been
watch
ing for weeks. The distinctive green neon sign above the door cast an eerie glow over the rain-soaked pavement. Quinn hesitated, her hand resting on the butt of her gun. Something didn't feel right. She pushed open the door and slipped inside, her eyes adjusting to the warm, smoky atmosphere. The Nest was a haven for the city's supernatural underbelly, and Quinn's presence
was bound
to draw attention. Herrera was nowhere to be seen, but Quinn spotted a familiar figure polishing a mug behind the bar – Silas, the owner. "Detective," Silas said, his voice low and even. "What brings you to my humble establishment?" Quinn flashed her badge. "I'm looking for Tomás Herrera. He just came in." Silas's expression didn't change, but Quinn detected a
flicker
of warning in his eyes. "Haven't seen him, Detective. Maybe he's in the... ah, restroom." Quinn's gaze swept the bar, taking in the patrons huddled over drinks and the shelves lined with dusty tomes. She knew Silas
was hiding
something, but she also knew she couldn't push him – not yet. As she waited, a tall, hooded figure emerged from the back room, a bookshelf sliding shut behind them.
The air was thick with
the scent of old books and something else – something earthy and decaying. "Excuse me," Quinn said, her eyes locked on the figure, but Silas intervened. "Sorry, Detective. That's a private meeting. Nothing to do with your... inquiry." Quinn's instincts
screamed that she
was missing something. Herrera had disappeared, and this hooded figure
seemed to
be the only lead. She weighed her options – follow the figure, or retreat and regroup. Her decision
was made
when the figure pushed back its hood, revealing a messy tangle of black hair. For an instant, their eyes met, and Quinn saw
a flicker of recognition
. Then, the figure turned and disappeared into the crowd, a small, bone-colored token clutched in their hand. Quinn's gut told her to follow. She pushed through the patrons, ignoring Silas's protests, and trailed the hooded figure out into the rain-soaked night. They led her on a wild goose chase, dodging market stalls and food carts in the deserted streets of Camden. Quinn's lungs burned, but she refused to give up. The figure finally stopped at the entrance to an abandoned Tube station, a rusty sign creaking in the wind. The Veil Market, Quinn realized, her mind racing. She had heard whispers of the underground market, but she had never been able to pin down its location. The figure produced the bone token and handed it to a burly man guarding the entrance. As Quinn hesitated, the guard's eyes met hers, and she felt a shiver run down her spine. This was uncharted territory, and she was alone. But she also knew that if she backed down now, she might never find the answers she sought. With a deep breath, Quinn followed the figure into the unknown, the darkness of the abandoned Tube station swallowing her whole.
created_at
"2026-02-17T02:21:06.112999+00:00"
2/17/2026, 2:21:06 AM
model
"meta-llama/llama-3.1-70b-instruct"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
428
moderation_latency
(null)
generation_time
29321
tokens_prompt
959
tokens_completion
1125
native_tokens_prompt
817
native_tokens_completion
960
native_tokens_completion_images
(null)
native_tokens_reasoning
0
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.0007108
router
(null)
provider_responses
0
id
"cmpl-50e0ae5cef3a46289aa7c0a4f6700961"
status
200
is_byok
false
latency
68
endpoint_id
"59c87462-40b6-4231-91cf-6d0f8f25e8b9"
provider_name
"DeepInfra"
model_permaslug
"meta-llama/llama-3.1-70b-instruct"
api_type
"completions"
id
"gen-1771294836-uAHplPgX948bU67Yy89e"
upstream_id
"cmpl-50e0ae5cef3a46289aa7c0a4f6700961"
total_cost
0.0007108
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
Evaluation details
Result
Evaluator
Details
Meta Data
100.0000%
Adverbs in dialogue tags
Target: ≤10% dialogue tags with adverbs
totalTags
2
adverbTagCount
0
adverbTags
(empty)
86.5410%
AI-ism adverb frequency
Target: <1% AI-ism adverbs (57 tracked)
wordCount
743
totalAiIsmAdverbs
1
found
0
adverb
"quickly"
count
1
highlights
0
"quickly"
100.0000%
AI-ism character names
Target: 0 AI-default names (17 tracked, −20% each)
codexExemptions
(empty)
found
(empty)
100.0000%
AI-ism location names
Target: 0 AI-default location names (33 tracked, −20% each)
codexExemptions
(empty)
found
(empty)
59.6231%
AI-ism word frequency
Target: <1% AI-ism words (140 tracked)
wordCount
743
totalAiIsms
3
found
0
word
"flicker"
count
2
1
word
"jaw clenched"
count
1
highlights
0
"flicker"
1
"jaw clenched"
33.3333%
Cliche density
Target: ≤1 cliche(s) per 800-word window
totalCliches
3
maxInWindow
3
found
0
label
"jaw/fists clenched"
count
1
1
label
"air was thick with"
count
1
2
label
"flicker of emotion"
count
1
highlights
0
"jaw clenched"
1
"The air was thick with"
2
"a flicker of recognition"
100.0000%
Emotion telling (show vs. tell)
Target: ≤3% sentences with emotion telling
emotionTells
0
narrationSentences
47
matches
(empty)
100.0000%
Filter word density
Target: ≤12% sentences with filter/hedge words
filterCount
1
hedgeCount
1
narrationSentences
47
filterMatches
0
"watch"
hedgeMatches
0
"seemed to"
100.0000%
Overuse of "that" (subordinate clause padding)
Target: ≤10% sentences with "that" clauses
thatCount
1
totalSentences
51
matches
0
"screamed that she"
84.9394%
Paragraph length variance
Target: CV ≥0.5 for paragraph word counts
totalParagraphs
20
mean
37.2
std
16.64
cv
0.447
sampleLengths
0
65
1
56
2
61
3
47
4
50
5
56
6
15
7
13
8
24
9
36
10
36
11
13
12
12
13
33
14
48
15
24
16
48
17
43
18
43
19
21
76.5957%
Passive voice overuse
Target: ≤5% passive sentences
passiveCount
4
totalSentences
47
matches
0
"been taken"
1
"was involved"
2
"was bound"
3
"was made"
100.0000%
Past progressive (was/were + -ing) overuse
Target: ≤10% past progressive verbs
pastProgressiveCount
1
totalVerbs
129
matches
0
"was hiding"
88.0000%
Purple prose (modifier overload)
Target: <4% adverbs, <2% -ly adverbs, no adj stacking
wordCount
702
adjectiveStacks
2
stackExamples
0
"short, curly brown hair"
1
"small, bone-colored token"
adverbCount
12
adverbRatio
0.017094017094017096
lyAdverbCount
7
lyAdverbRatio
0.009971509971509971
100.0000%
Repeated phrase echo
Target: ≤20% sentences with echoes (window: 2)
totalSentences
51
echoCount
0
echoWords
(empty)
66.8977%
Sentence length variance
Target: CV ≥0.4 for sentence word counts
totalSentences
51
mean
14.59
std
4.63
cv
0.317
sampleLengths
0
15
1
14
2
17
3
19
4
24
5
17
6
15
7
14
8
15
9
18
10
14
11
16
12
13
13
18
14
19
15
16
16
11
17
4
18
16
19
18
20
22
21
8
22
7
23
4
24
9
25
14
26
10
27
19
28
17
29
18
30
18
31
13
32
12
33
8
34
13
35
12
36
18
37
13
38
17
39
6
40
18
41
20
42
9
43
19
44
8
45
19
46
16
47
17
48
8
49
18
76.4706%
Sentence opener variety
Target: ≥60% unique sentence openers
consecutiveRepeats
1
diversityRatio
0.47058823529411764
totalSentences
51
uniqueOpeners
24
50.0000%
Dialogue tag variety (said vs. fancy)
Target: ≤30% fancy dialogue tags
totalTags
2
fancyCount
1
fancyTags
0
"Quinn said, her eyes locked on the figure, but Silas intervened (say, lock intervene)"
83.6706%