Run: 2026-02-19-01KHT00218PB3EK9NDF5HVJ1RT

Run Details

Model:
DeepSeek V3 (2025-03-24)
Test:
Bad Writing Habits
Scenario:
Mystery: examining a crime scene
Duration:
42.2s
Cost:
$0.000824
Provider:
DeepInfra

Evaluation summary

Score Evaluator
Details Link
40.00% Adverbs in dialogue tags (details)
71.70% AI-ism adverb frequency (details)
100.00% AI-ism character names (details)
100.00% AI-ism location names (details)
33.96% AI-ism word frequency (details)
100.00% Cliché density (details)
100.00% Emotion telling (show vs. tell) (details)
100.00% Filter word density (details)
100.00% Gibberish response detection (details)
100.00% Markdown formatting overuse (details)
100.00% Missing dialogue indicators (quotation marks) (details)
0.00% Name drop frequency (details)
53.85% Narrator intent-glossing (details)
100.00% "Not X but Y" pattern overuse (details)
100.00% Overuse of "that" (subordinate clause padding) (details)
100.00% Paragraph length variance (details)
90.94% Passive voice overuse (details)
100.00% Past progressive (was/were + -ing) overuse (details)
0.00% Em-dash & semicolon overuse (details)
100.00% Purple prose (modifier overload) (details)
100.00% Repeated phrase echo (details)
100.00% Sentence length variance (details)
69.79% Sentence opener variety (details)
0.00% Adverb-first sentence starts (details)
100.00% Pronoun-first sentence starts (details)
26.67% Subject-first sentence starts (details)
0.00% Subordinate conjunction sentence starts (details)
83.33% Technical jargon density (details)
75.00% Useless dialogue additions (details)
100.00% Dialogue tag variety (said vs. fancy) (details)
74.8413% Overall

Model input

System
You are an expert fiction writer. Always keep the following rules in mind: - Write in past tense and use General English spelling, grammar, and colloquialisms/slang. - Write in active voice - Always follow the "show, don't tell" principle. - Avoid adverbs and cliches and overused/commonly used phrases. Aim for fresh and original descriptions. - Convey events and story through dialogue. - Mix short, punchy sentences with long, descriptive ones. Drop fill words to add variety. - Skip "he/she said said" dialogue tags and convey people's actions or face expressions through their speech - Avoid mushy dialog and descriptions, have dialogue always continue the action, never stall or add unnecessary fluff. Vary the descriptions to not repeat yourself. - Put dialogue on its own paragraph to separate scene and action. - Reduce indicators of uncertainty like "trying" or "maybe" When writing text: - NEVER conclude the scene on your own, follow the beat instructions very closely. - NEVER end with foreshadowing. - NEVER write further than what I prompt you with. - AVOID imagining possible endings, NEVER deviate from the instructions. - STOP EARLY if the continuation contains what was required in the instructions. You do not need to fill out the full amount of words possible.
User/Human
Take into account the following glossary of characters/locations/items/lore... when writing your response: <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="4414" name="Eva Kowalski" aliases="Eva, Evie" pronouns="she/her" species="Human" storyRole="Best Friend"> <eyeColor>green</eyeColor> <hair>curly red</hair> <height>5'4"</height> <age>26</age> <glasses>round glasses</glasses> <satchel>worn leather satchel full of books</satchel> <freckles>freckled complexion</freckles> <nervousHabit>tucking hair behind her left ear</nervousHabit> <description>Aurora's childhood friend and an occult researcher.</description> <backstory>Moved to London two years before Aurora and works as a research assistant at the British Museum's restricted archives. She holds a master's degree in Ancient History from Oxford.</backstory> </character> <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> <item id="2004" name="Veil Compass" aliases="the Compass, shadow compass" type="Tool" origin="The Veil Market"> <casingMaterial>small brass compass</casingMaterial> <faceMarkings>face etched with protective sigils</faceMarkings> <needle>points toward the nearest supernatural rift or portal</needle> <patina>casing has a patina of verdigris</patina> <creator>crafted by a Shade artisan</creator> <description>A compass attuned to supernatural energy.</description> </item> </codex> Continue the story using the following instructions: <instructions> Write a fully fleshed out scene of at least 1000 words in the mystery genre. A detective arrives at a crime scene in an unusual location. Something is wrong — the evidence doesn't add up. A colleague is present, offering a different interpretation. Through careful observation and deduction, the detective begins to see what others have missed. 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. </instructions>

Model response

AI
The abandoned Tube station smelled of damp concrete and something metallic underneath—blood, maybe, or rust from the old tracks. Detective Harlow Quinn stepped over the yellow police tape, her boots crunching on broken glass. The flickering emergency lights cast long shadows across the platform, turning the graffiti into writhing shapes. "Quinn." Eva Kowalski stood near the edge of the platform, her freckled face pale under the harsh light. She clutched her worn leather satchel like a shield, round glasses reflecting the dim glow. "You shouldn’t be here." Quinn kept her voice low, scanning the scene. Two uniforms stood by a tarp-covered shape near the tracks. The air hummed with tension. "I was already here," Eva said, tucking a curl behind her ear. "Researching. Then the—the body showed up." Quinn moved past her, crouching beside the tarp. She lifted the edge. The corpse was male, mid-thirties, dressed in a suit that had seen better days. His throat was slit clean, but the blood pooling beneath him was wrong—too dark, almost black. The edges of the wound shimmered faintly, like oil under light. "Knife wound?" one of the uniforms offered. Quinn didn’t answer. She pulled out her flashlight, angling the beam. The man’s left sleeve was rolled up, revealing a tattoo—a series of interlocking circles. Eva sucked in a breath. "That’s a Veil Mark." Quinn glanced up. "Explain." "It’s—it’s a sigil. Means he was a trader here. The Market moves every full moon, but some regulars get marked so they can find it." Eva’s fingers twitched toward her satchel. "This wasn’t just a mugging." Quinn stood, wiping her hands on her coat. "No kidding." She paced the platform, studying the scene. No signs of struggle. No defensive wounds. The man had been killed quickly, efficiently. But the blood— "Where’s his token?" Eva murmured. Quinn turned. "What?" "Bone token. You need one to get in. Traders always carry them." Eva knelt, gingerly checking the man’s pockets. "It’s gone." Quinn exhaled sharply. "So someone took it. They wanted access." "Or they didn’t want him to have it anymore." Eva’s voice wavered. A uniform cleared his throat. "Detective, we’ve got a witness. Says he saw the vic arguing with someone in a long coat before it happened." Quinn’s pulse spiked. "Description?" "Tall. Hooded. Didn’t see a face." Eva made a small noise. Quinn shot her a look. "You know something." Eva hesitated, then dug into her satchel. She pulled out a small brass compass, its face etched with strange symbols. The needle spun wildly before settling, pointing toward the tunnel’s mouth. "That’s not just a compass," Quinn said. Eva swallowed. "It points to rifts. Portals. Right now, it’s saying there’s one down there." Quinn stared into the tunnel’s darkness. The air smelled like ozone now, sharp and electric. "This wasn’t a robbery," she said slowly. "This was a message. Someone killed him to send a warning—or to keep him quiet." Eva’s fingers tightened around the compass. "We should go. Before they come back." Quinn didn’t move. The pieces clicked together—the missing token, the sigil, the unnatural blood. Someone was playing a game. And she intended to win.
created_at"2026-02-19T03:44:45.712Z" 2/19/2026, 3:44:45 AM
model"deepseek/deepseek-chat-v3-0324"
app_id182717
external_user(null)
streamedtrue
cancelledfalse
latency1081
moderation_latency(null)
generation_time41424
tokens_prompt1149
tokens_completion884
native_tokens_prompt1032
native_tokens_completion808
native_tokens_completion_images(null)
native_tokens_reasoning0
native_tokens_cached64
num_media_prompt(null)
num_input_audio_prompt(null)
num_media_completion0
num_search_results(null)
origin"https://ncbench.com/"
is_byokfalse
finish_reason"stop"
native_finish_reason"stop"
usage0.0008244
router(null)
provider_responses
0
endpoint_id"820376cb-f110-4d56-ab52-5bd6ca269420"
id"chatcmpl-ROQ7vPEr5ENRnRp6J6e6T5vd"
is_byokfalse
latency58
model_permaslug"deepseek/deepseek-chat-v3-0324"
provider_name"DeepInfra"
status200
api_type"completions"
id"gen-1771472685-vHAX2OowKqurdE0VzonH"
upstream_id"chatcmpl-ROQ7vPEr5ENRnRp6J6e6T5vd"
total_cost0.0008244
cache_discount0.00000416
upstream_inference_cost0
provider_name"DeepInfra"

Evaluation details

Result Evaluator Details Meta Data
40.00% Adverbs in dialogue tags
Target: ≤10% dialogue tags with adverbs
totalTags10
adverbTagCount2
adverbTags
0"Eva knelt gingerly [gingerly]"
1"she said slowly [slowly]"
dialogueSentences25
tagDensity0.4
leniency0.8
rawRatio0.2
effectiveRatio0.16
71.70% AI-ism adverb frequency
Target: <2% AI-ism adverbs (58 tracked)
wordCount530
totalAiIsmAdverbs3
found
0
adverb"quickly"
count1
1
adverb"sharply"
count1
2
adverb"slowly"
count1
highlights
0"quickly"
1"sharply"
2"slowly"
100.00% AI-ism character names
Target: 0 AI-default names (17 tracked, −20% each)
codexExemptions(empty)
found(empty)
100.00% AI-ism location names
Target: 0 AI-default location names (33 tracked, −20% each)
codexExemptions(empty)
found(empty)
33.96% AI-ism word frequency
Target: <2% AI-ism words (290 tracked)
wordCount530
totalAiIsms7
found
0
word"scanning"
count1
1
word"tension"
count1
2
word"shimmered"
count1
3
word"wavered"
count1
4
word"pulse"
count1
5
word"etched"
count1
6
word"electric"
count1
highlights
0"scanning"
1"tension"
2"shimmered"
3"wavered"
4"pulse"
5"etched"
6"electric"
100.00% Cliché density
Target: ≤1 cliche(s) per 800-word window
totalCliches0
maxInWindow0
found(empty)
highlights(empty)
100.00% Emotion telling (show vs. tell)
Target: ≤3% sentences with emotion telling
emotionTells0
narrationSentences49
matches(empty)
100.00% Filter word density
Target: ≤3% sentences with filter/hedge words
filterCount0
hedgeCount1
narrationSentences49
filterMatches(empty)
hedgeMatches
0"tended to"
100.00% Gibberish response detection
Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words)
analyzedSentences64
gibberishSentences0
adjustedGibberishSentences0
longSentenceCount0
runOnParagraphCount0
giantParagraphCount0
wordSaladCount0
repetitionLoopCount0
controlTokenCount0
maxSentenceWordsSeen31
ratio0
matches(empty)
100.00% Markdown formatting overuse
Target: ≤5% words in markdown formatting
markdownSpans0
markdownWords0
totalWords522
ratio0
matches(empty)
100.00% Missing dialogue indicators (quotation marks)
Target: ≤10% speech attributions without quotation marks
totalAttributions8
unquotedAttributions0
matches(empty)
0.00% Name drop frequency
Target: ≤1.0 per-name mentions per 100 words
totalMentions28
wordCount360
uniqueNames6
maxNameDensity3.61
worstName"Quinn"
maxWindowNameDensity5
worstWindowName"Quinn"
discoveredNames
Tube1
Harlow1
Quinn13
Kowalski1
Two1
Eva11
persons
0"Harlow"
1"Quinn"
2"Kowalski"
3"Eva"
places(empty)
globalScore0
windowScore0
53.85% Narrator intent-glossing
Target: ≤2% narration sentences with intent-glossing patterns
analyzedSentences26
glossingSentenceCount1
matches
0"smelled like ozone now, sharp and electric"
100.00% "Not X but Y" pattern overuse
Target: ≤1 "not X but Y" per 1000 words
totalMatches0
per1kWords0
wordCount522
matches(empty)
100.00% Overuse of "that" (subordinate clause padding)
Target: ≤2% sentences with "that" clauses
thatCount0
totalSentences64
matches(empty)
100.00% Paragraph length variance
Target: CV ≥0.5 for paragraph word counts
totalParagraphs30
mean17.4
std12
cv0.69
sampleLengths
050
133
227
318
412
541
67
725
89
94
1036
1110
1224
135
143
1521
1610
1712
1825
194
206
2110
223
2331
247
2515
2615
2722
2813
2924
90.94% Passive voice overuse
Target: ≤2% passive sentences
passiveCount2
totalSentences49
matches
0"was rolled"
1"been killed"
100.00% Past progressive (was/were + -ing) overuse
Target: ≤2% past progressive verbs
pastProgressiveCount1
totalVerbs71
matches
0"was playing"
0.00% Em-dash & semicolon overuse
Target: ≤2% sentences with em-dashes/semicolons
emDashCount5
semicolonCount0
flaggedSentences5
totalSentences64
ratio0.078
matches
0"The abandoned Tube station smelled of damp concrete and something metallic underneath—blood, maybe, or rust from the old tracks."
1"His throat was slit clean, but the blood pooling beneath him was wrong—too dark, almost black."
2"The man’s left sleeve was rolled up, revealing a tattoo—a series of interlocking circles."
3"But the blood—"
4"The pieces clicked together—the missing token, the sigil, the unnatural blood."
100.00% Purple prose (modifier overload)
Target: <4% adverbs, <2% -ly adverbs, no adj stacking
wordCount366
adjectiveStacks0
stackExamples(empty)
adverbCount14
adverbRatio0.03825136612021858
lyAdverbCount7
lyAdverbRatio0.01912568306010929
100.00% Repeated phrase echo
Target: ≤20% sentences with echoes (window: 2)
totalSentences64
echoCount0
echoWords(empty)
100.00% Sentence length variance
Target: CV ≥0.4 for sentence word counts
totalSentences64
mean8.16
std5.77
cv0.708
sampleLengths
019
115
216
318
415
512
610
75
812
96
108
114
1214
1316
1411
157
163
178
1814
195
204
213
221
2331
245
258
262
277
284
293
307
313
325
332
341
3519
362
373
387
3912
405
4120
423
431
446
455
465
473
487
4913
69.79% Sentence opener variety
Target: ≥60% unique sentence openers
consecutiveRepeats2
diversityRatio0.4375
totalSentences64
uniqueOpeners28
0.00% Adverb-first sentence starts
Target: ≥3% sentences starting with an adverb
adverbCount0
totalSentences45
matches(empty)
ratio0
100.00% Pronoun-first sentence starts
Target: ≤30% sentences starting with a pronoun
pronounCount7
totalSentences45
matches
0"She clutched her worn leather"
1"She lifted the edge."
2"His throat was slit clean,"
3"She pulled out her flashlight,"
4"She paced the platform, studying"
5"She pulled out a small"
6"she said slowly"
ratio0.156
26.67% Subject-first sentence starts
Target: ≤72% sentences starting with a subject
subjectCount39
totalSentences45
matches
0"The abandoned Tube station smelled"
1"Detective Harlow Quinn stepped over"
2"The flickering emergency lights cast"
3"Eva Kowalski stood near the"
4"She clutched her worn leather"
5"Quinn kept her voice low,"
6"The air hummed with tension."
7"Eva said, tucking a curl"
8"Quinn moved past her, crouching"
9"She lifted the edge."
10"The corpse was male, mid-thirties,"
11"His throat was slit clean,"
12"The edges of the wound"
13"Quinn didn’t answer."
14"She pulled out her flashlight,"
15"The man’s left sleeve was"
16"Eva sucked in a breath."
17"Quinn glanced up."
18"Eva’s fingers twitched toward her"
19"Quinn stood, wiping her hands"
ratio0.867
0.00% Subordinate conjunction sentence starts
Target: ≥2% sentences starting with a subordinating conjunction
subConjCount0
totalSentences45
matches(empty)
ratio0
83.33% Technical jargon density
Target: ≤6% sentences with technical-jargon patterns
analyzedSentences12
technicalSentenceCount1
matches
0"The corpse was male, mid-thirties, dressed in a suit that had seen better days."
75.00% Useless dialogue additions
Target: ≤5% dialogue tags with trailing filler fragments
totalTags10
uselessAdditionCount1
matches
0"Eva Kowalski stood, her freckled face pale under the harsh light"
100.00% Dialogue tag variety (said vs. fancy)
Target: ≤10% fancy dialogue tags
totalTags4
fancyCount1
fancyTags
0"Eva murmured (murmur)"
dialogueSentences25
tagDensity0.16
leniency0.32
rawRatio0.25
effectiveRatio0.08
74.8413%