NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HCC6DQ5RD28Q2NCKMF9CW0
Run Details
Model:
Qwen 3.7 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
1m 5s
Cost:
$0.001428
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
86.38%
Matches word count
(details)
81.87%
Dialogue to Total Word Ratio
(details)
84.1267%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% words of dialogue between two thieves named Molly and Bill who are planning to break into a children's party to steal all the presents. Only include the prose of the scene without any markdown or commentary.
Model response
Messages
Meta
AI
The heat radiating from the inflatable dragon, shaped like a vulgar orange beast, made Molly wipe her brow with a grease-stained rag. She grimaced as a droplet of sweat traced a path through the grime on her cheek, mixing with the invisible dust motes that floated in the stagnant air between the bouncy castle and the rotting porch swing. Bill adjusted his goggles, which were fogged by the humidity of eighty screaming toddlers nearby, their high-pitched shrieks vibrating through the wood of the garage walls like angry hornets. He pointed a laser thermometer at the house where wrapping paper piled over a box wrapped in neon green dinosaur paper, the red dot trembling violently over the cardboard seam as if sensing the weight of the forbidden technology inside. His hands shook as he calibrated the dial, sweating profusely not from the oppressive summer heat, but from the crushing weight of the moral decay they were about to enact upon the unsuspecting residents of 42 Maple Drive. "Target acquired," Bill whispered. "The Master Bedroom is upstairs, left wing. That's where the big stuff is. There's a gaming console worth more than my car, and possibly a unicycle." Molly checked the stopwatch taped over her band-aid, the plastic strap cutting into her skin as she twisted the crown to reset the timer. The strategy relied entirely on the host's biological failure; human digestion was a predictable weapon when paired with refined sugars and ethanol. "We have forty-five minutes before the cake cuts. Parents will be too drunk on cheap wine to notice two silhouettes slipping out the window with duffel bags full of Joy-Cons." Bill shrank back against the garage siding, clutching his grappling hook. "And the dogs? What about Fluffy? I read the dossier. Fluffy bites ankles." "Fluffy is asleep after eating that tub of kibble I bribed her with during stakeout number three," Molly assured him. "Besides, we have the laser cutters. We'll slice through the blinds without waking a soul. Then it's zip-line to the drainpipe, traverse the trellis, and drop into the minivan." Bill narrowed his eyes behind the foggy lenses. "You mean the getaway vehicle smells like old gym socks and regret?" "It smells like victory, Bill. Can you stop whining? We're professionals. We stole the Mayor's toupee last Tuesday." "Stealing presents violates the code. Santa has jurisdiction here." "Santa operates on a global logistics network with tax-exempt status," Molly countered sharply. "We operate on cash and adrenaline. Look at that stack of boxes. You see loot. Pure, unadulterated loot. The LEGO Death Star alone could fund our retirement in Belize." Bill licked his lips. "With the all-inclusive resort?" "With the all-inclusive resort. And a hammock made of pure silk." Molly extended a hand. "Ready to ruin a seven-year-old's life for the money?" Bill looked at the house, then gripped Molly's hand firmly. "Lead the way, partner. But if I hear jingle bells, I'm blaming you."
created_at
"2026-09-02T15:39:05.025Z"
9/2/2026, 3:39:05 PM
model
"qwen/qwen3.7-flash-20260727"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
372
moderation_latency
(null)
generation_time
64968
tokens_prompt
75
tokens_completion
9680
native_tokens_prompt
76
native_tokens_completion
10967
native_tokens_completion_images
(null)
native_tokens_reasoning
10309
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
num_fetches
(null)
web_search_engine
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
service_tier
(null)
usage
0.00142799
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788363545-lXZUqnBkum4XNfoODbhL"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788363545-vPppfrUuSMBqVkRaIH7K"
upstream_id
"chatcmpl-4a829588-a957-9d0f-8392-9d7686006a67"
provider_responses
0
endpoint_id
"5a9a0ee6-874a-4d03-92be-e5094c0df4e7"
id
"chatcmpl-4a829588-a957-9d0f-8392-9d7686006a67"
is_byok
false
latency
372
model_permaslug
"qwen/qwen3.7-flash-20260727"
provider_name
"Alibaba"
status
200
total_cost
0.00142799
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Alibaba"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
86.38%
Matches word count
n/a
neededClean
false
words
489
81.87%
Dialogue to Total Word Ratio
Ratio: 48.00%, Deviation: 2.00%
neededClean
false
wordsTotal
500
wordsDialogue
240
84.1267%