Run: 2026-10-13-01M4TB4WA5V6J0GCN57KE5YSVG

Run Details

Duration:
2m 10s
Cost:
$0.106791
Provider:
Fireworks

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
By the third flight of stairs, Quinn's torch had gone the colour of weak tea. She tapped it against her palm. The beam rallied, then shrank to a thread. Below, in the ticket hall, Doyle had planted his own torch upright on the floor, and its cone lit the ceiling tiles from beneath — the green of old bottle glass, flaked with ninety years of damp. Hawley Crescent station had closed in 1934. The dark down here had had time to settle. "Guv." Doyle worked his gum, hands in the pockets of his suit jacket. "You're going to love this one." "Give it to me." "Anonymous call, ten past four. Woman's voice. Said there's a body on the platform at Hawley Crescent, then hung up before the operator got a name. Uniform came in through the maintenance door on the alley. Street gate's rusted solid — nobody's opened it since the war." "And the caller knew he was here how?" "That's the bit you'll love." He nodded down the passage. "Come and see." The body lay at the centre of the platform, on its back, arms arranged along its sides. Male, mid-thirties. A wool coat that cost real money, clean shoes, manicured nails. Wrong, all of it, for a squat or a squatters' party. Dr Banerji crouched beside him with her kit open. "Morning, Detective. No wounds. No puncture sites, no ligature marks, no petechiae. His heart stopped. That's what I have until the post-mortem." "Time?" "Between eleven and two, going by temperature and rigor. Lividity's fixed on his back. He died where he lies — nobody moved him after." Quinn circled the body at two metres, torch raised. The dimming beam caught pools of candle wax at regular intervals along both platforms, and beneath the dust, rectangles scuffed clean — rows of them, in lanes, the length of the station. Chalk lines marked the corners of each. "Rave." Doyle spread his hands. "Or a buy. Dealers set up tables, punters come down, something goes wrong for this one. They bolt, he stays. Happens." "Count the cans, Aaron." "What?" "Cans. Bottles. Fag ends. Needles. Vomit in the corners. Find me one." She walked the nearest lane, slow, reading the floor. "A hundred people stood down here for hours. I've policed raves in warehouses, in fields, in a drained lido. They all leave a skin. This lot left nothing. Not a crisp packet." "Maybe they were tidy." "Dealers don't chalk their pitches and light them with candles." A SOCO named Lena looked up from the threshold of the ticket hall, swab in hand. "Detective. The lines across the doorways. I've tested twice. It's salt. Just salt." "Salt at every entrance," Quinn said. "Ritual lot, then." Doyle shrugged. "Occult-themed party. Camden's full of them." Quinn's torch guttered. She checked her watch out of habit — the leather strap had gone soft and dark with years of wear, and the mechanical hands ticked on, unbothered. Doyle saw the gesture and pulled out his own phone. "Dead. Thirty-four per cent when I came down. Yours?" "Watch is wound, not wired." She turned the thought over. "Whatever's flattening batteries down here has no taste for springs." "Cheap torches, guv." She walked to the stairs instead of answering. The dust on the steps held prints in plenty — dozens of sets, overlapping, all sizes. She crouched. Turned. Looked up the flight, then down, and felt the shape of it before the words came. "Lena. Photograph the stairs, every tread. Then tell me how many prints go up." Ten minutes later Lena came back with her camera screen held out and her mouth set in a line. "Down only. I count ninety, maybe a hundred distinct treads descending. Nothing ascends. Not one." "Service tunnel." Doyle was already moving. "Track level. They went out along the line." They followed him along the trackbed — a dry canal of soot where the rails had been lifted decades before — to the tunnel mouth. Victorian brick sealed it, black with old smoke, the mortar crusted. Cobwebs sagged across an inspection hatch, grey and unbroken. "Bricked up in the fifties," Lena said. "Soot's intact. Nothing's gone through there since." Doyle's jaw worked the gum harder. Back on the platform, Quinn knelt by the body and lifted the cord around the dead man's neck. A disc hung from it — bone, thumb-width, polished by handling, carved with a mark: a keyhole inside a spiral. Three years since Morris. His notebook had carried a page of that same mark, copied over and over in his cramped hand, the page he never explained to anyone. She'd memorised it without meaning to, the way the mind keeps what a case leaves open. She set the disc down and stood before the memory could take hold. In his coat pocket she found the compass. Brass, small enough to vanish in a fist, its casing furred with verdigris, the face etched with sigils too fine for machine work. She held it level. The needle pointed down the platform. "North's that way." Doyle pointed the opposite direction. "I know where north is." She turned a full circle, slow. The needle swung with her and settled, every time, on the bricked arch at the platform's end. She walked it back to the body. Over the dead man, the needle lost its purpose and wandered in loose circles. Ten steps away, it snapped back to the arch. "Magnet in the brickwork. Old ironwork in the walls. It's a prop, guv. A toy from one of the stalls." "Then it'll point at any iron." She held it over the trackbed, over Banerji's steel kit case, over Doyle's belt buckle. The needle ignored all of it. It wanted the wall. Lena called from the far lane, voice pitched careful, the way people speak when the floor has just shifted. "Detective. You should see the overlay." The photographs told it in sequence. The scuffed lanes between the pitches crossed and re-crossed — and where the lanes reached the centre of the platform, they bent. A clean arc around one fixed point, trodden by dozens of feet, over and over, through the whole shape of the night. "The lanes cut across the earlier prints," Lena said. "The crowd walked this way for hours. They walked around something. Around him." Doyle stared at the screen, then at the body, then back. "He died after they left. Midnight, Banerji said. Market packed up at eleven." "Lividity says he never moved. He lay here, dying or dead, at the centre of a working market, and ninety people stepped around him like he was a pillar box." Quinn let that sit. "Nobody helped him. Nobody ran. The stalls traded on, and when it was over, they took their tables, their stock, their litter, every trace of themselves — and left him." "All right." Doyle's voice had lost the gum-chewing ease. "Say it was a market. Ninety-odd people come down those stairs. Nobody goes back up. The tunnel's bricked. The gate's rusted shut. Where did they go?" Quinn walked to the arch. The brick stood solid, soot-black, undisturbed, the cobwebs whole. She raised the compass. The needle held true on the wall, steady as a blade, and would not be moved. "They didn't go up, Aaron. Ninety people. Check the tread count again if you like." "People don't walk through walls." "No." The sigils on the compass face caught the last of her torchlight, and the needle did not waver. "So between the bottom of those stairs and this wall, there's a door. Find it."
created_at"2026-10-13T09:57:52.079Z" 10/13/2026, 9:57:52 AM
model"fireworks/ember-1-20260923"
app_id182717
external_user(null)
streamedtrue
cancelledfalse
latency2906
moderation_latency(null)
generation_time130198
tokens_prompt1112
tokens_completion7673
native_tokens_prompt1110
native_tokens_completion7097
native_tokens_completion_images(null)
native_tokens_reasoning5359
native_tokens_cached1109
num_media_prompt(null)
num_input_audio_prompt(null)
num_media_completion0
num_search_results(null)
num_fetches(null)
web_search_engine(null)
origin"https://ncbench.com/"
is_byokfalse
finish_reason"stop"
native_finish_reason"stop"
service_tier(null)
usage0.1067907
router(null)
user_agent"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer(null)
request_id"req-1791885472-icX9MPa0PAYth4gzoVd0"
session_id(null)
preset_id(null)
api_type"completions"
id"gen-1791885472-6LmyGUUGRQ7Jp0KQ0ii7"
upstream_id"chatcmpl-b4c6418b5d884969ba13ad7834ad2d52"
provider_responses
0
endpoint_id"b89a2f05-4051-4446-a9b0-7bc8ae7b7c90"
id"chatcmpl-b4c6418b5d884969ba13ad7834ad2d52"
is_byokfalse
latency2906
model_permaslug"fireworks/ember-1-20260923"
provider_name"Fireworks"
status200
total_cost0.1067907
cache_discount0.0029943
upstream_inference_cost0
provider_name"Fireworks"
response_cache_source_id(null)
data_region"global"
workspace_id"97e315e5-d303-487d-83c1-83180e8a13d4"

Evaluation details

Result Evaluator Details Meta Data
100.00% Adverbs in dialogue tags
Target: ≤10% dialogue tags with adverbs
totalTags14
adverbTagCount0
adverbTags(empty)
dialogueSentences48
tagDensity0.292
leniency0.583
rawRatio0
effectiveRatio0
100.00% AI-ism adverb frequency
Target: <2% AI-ism adverbs (58 tracked)
wordCount1255
totalAiIsmAdverbs0
found(empty)
highlights(empty)
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)
92.03% AI-ism word frequency
Target: <2% AI-ism words (290 tracked)
wordCount1255
totalAiIsms2
found
0
word"mechanical"
count1
1
word"etched"
count1
highlights
0"mechanical"
1"etched"
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
narrationSentences69
matches(empty)
100.00% Filter word density
Target: ≤3% sentences with filter/hedge words
filterCount1
hedgeCount0
narrationSentences69
filterMatches
0"watch"
hedgeMatches(empty)
100.00% Gibberish response detection
Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words)
analyzedSentences101
gibberishSentences0
adjustedGibberishSentences0
longSentenceCount0
runOnParagraphCount0
giantParagraphCount0
wordSaladCount0
repetitionLoopCount0
controlTokenCount0
repeatedSegmentCount0
maxSentenceWordsSeen46
ratio0
matches(empty)
100.00% Markdown formatting overuse
Target: ≤5% words in markdown formatting
markdownSpans0
markdownWords0
totalWords1266
ratio0
matches(empty)
100.00% Missing dialogue indicators (quotation marks)
Target: ≤10% speech attributions without quotation marks
totalAttributions10
unquotedAttributions1
matches
0"Ten steps away, it snapped back to the arch."
77.34% Name drop frequency
Target: ≤1.0 per-name mentions per 100 words
totalMentions28
wordCount757
uniqueNames7
maxNameDensity1.45
worstName"Doyle"
maxWindowNameDensity2.5
worstWindowName"Doyle"
discoveredNames
Quinn7
Doyle11
Crescent1
Dr1
Banerji2
Lena5
Morris1
persons
0"Quinn"
1"Doyle"
2"Dr"
3"Banerji"
4"Lena"
5"Morris"
places
0"Crescent"
globalScore0.773
windowScore0.833
100.00% Narrator intent-glossing
Target: ≤2% narration sentences with intent-glossing patterns
analyzedSentences50
glossingSentenceCount0
matches(empty)
100.00% "Not X but Y" pattern overuse
Target: ≤1 "not X but Y" per 1000 words
totalMatches0
per1kWords0
wordCount1266
matches(empty)
100.00% Overuse of "that" (subordinate clause padding)
Target: ≤2% sentences with "that" clauses
thatCount1
totalSentences101
matches
0"let that sit"
100.00% Paragraph length variance
Target: CV ≥0.5 for paragraph word counts
totalParagraphs53
mean23.89
std18.1
cv0.758
sampleLengths
015
167
219
34
447
58
613
750
822
91
1024
1148
1226
134
141
1553
164
1710
1816
1913
206
2111
2240
239
2420
253
2643
2714
2819
2915
3014
3145
3214
336
3438
3558
3641
378
3858
3920
4031
4119
426
4350
4422
4511
4613
4764
4835
4934
100.00% Passive voice overuse
Target: ≤2% passive sentences
passiveCount1
totalSentences69
matches
0"been lifted"
100.00% Past progressive (was/were + -ing) overuse
Target: ≤2% past progressive verbs
pastProgressiveCount1
totalVerbs120
matches
0"was already moving"
0.00% Em-dash & semicolon overuse
Target: ≤2% sentences with em-dashes/semicolons
emDashCount8
semicolonCount0
flaggedSentences7
totalSentences101
ratio0.069
matches
0"Below, in the ticket hall, Doyle had planted his own torch upright on the floor, and its cone lit the ceiling tiles from beneath — the green of old bottle glass, flaked with ninety years of damp."
1"The dimming beam caught pools of candle wax at regular intervals along both platforms, and beneath the dust, rectangles scuffed clean — rows of them, in lanes, the length of the station."
2"She checked her watch out of habit — the leather strap had gone soft and dark with years of wear, and the mechanical hands ticked on, unbothered."
3"The dust on the steps held prints in plenty — dozens of sets, overlapping, all sizes."
4"They followed him along the trackbed — a dry canal of soot where the rails had been lifted decades before — to the tunnel mouth."
5"A disc hung from it — bone, thumb-width, polished by handling, carved with a mark: a keyhole inside a spiral."
6"The scuffed lanes between the pitches crossed and re-crossed — and where the lanes reached the centre of the platform, they bent."
100.00% Purple prose (modifier overload)
Target: <4% adverbs, <2% -ly adverbs, no adj stacking
wordCount495
adjectiveStacks0
stackExamples(empty)
adverbCount10
adverbRatio0.020202020202020204
lyAdverbCount1
lyAdverbRatio0.00202020202020202
100.00% Repeated phrase echo
Target: ≤20% sentences with echoes (window: 2)
totalSentences101
echoCount0
echoWords(empty)
100.00% Sentence length variance
Target: CV ≥0.4 for sentence word counts
totalSentences101
mean12.53
std8.69
cv0.693
sampleLengths
015
16
28
337
47
59
613
76
84
947
108
1110
123
1317
142
1511
1611
179
1822
191
2024
219
2232
237
245
2521
264
271
2821
2932
304
3110
3216
3313
346
355
366
373
3827
3910
409
4110
4210
433
448
4516
462
471
4816
4914
97.03% Sentence opener variety
Target: ≥60% unique sentence openers
consecutiveRepeats3
diversityRatio0.6732673267326733
totalSentences101
uniqueOpeners68
0.00% Adverb-first sentence starts
Target: ≥3% sentences starting with an adverb
adverbCount0
totalSentences62
matches(empty)
ratio0
100.00% Pronoun-first sentence starts
Target: ≤30% sentences starting with a pronoun
pronounCount16
totalSentences62
matches
0"She tapped it against her"
1"He nodded down the passage"
2"She walked the nearest lane,"
3"She checked her watch out"
4"She turned the thought over"
5"She walked to the stairs"
6"They followed him along the"
7"His notebook had carried a"
8"She'd memorised it without meaning"
9"She set the disc down"
10"She held it level."
11"She turned a full circle,"
12"She walked it back to"
13"She held it over the"
14"It wanted the wall."
15"She raised the compass."
ratio0.258
40.65% Subject-first sentence starts
Target: ≤72% sentences starting with a subject
subjectCount52
totalSentences62
matches
0"She tapped it against her"
1"The beam rallied, then shrank"
2"Hawley Crescent station had closed"
3"The dark down here had"
4"Doyle worked his gum, hands"
5"He nodded down the passage"
6"The body lay at the"
7"A wool coat that cost"
8"Dr Banerji crouched beside him"
9"Quinn circled the body at"
10"The dimming beam caught pools"
11"Chalk lines marked the corners"
12"Doyle spread his hands"
13"She walked the nearest lane,"
14"A SOCO named Lena looked"
15"Quinn's torch guttered."
16"She checked her watch out"
17"Doyle saw the gesture and"
18"She turned the thought over"
19"She walked to the stairs"
ratio0.839
0.00% Subordinate conjunction sentence starts
Target: ≥2% sentences starting with a subordinating conjunction
subConjCount0
totalSentences62
matches(empty)
ratio0
100.00% Technical jargon density
Target: ≤6% sentences with technical-jargon patterns
analyzedSentences27
technicalSentenceCount0
matches(empty)
100.00% Useless dialogue additions
Target: ≤5% dialogue tags with trailing filler fragments
totalTags14
uselessAdditionCount0
matches(empty)
100.00% Dialogue tag variety (said vs. fancy)
Target: ≤10% fancy dialogue tags
totalTags3
fancyCount0
fancyTags(empty)
dialogueSentences48
tagDensity0.063
leniency0.125
rawRatio0
effectiveRatio0
86.9017%