NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HAJ2M7HTEZCHCH9RDTDMNT
Run Details
Model:
Qwen 3.7 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
50.6s
Cost:
$0.000966
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.99%
Dialogue to Total Word Ratio
(details)
0.4961%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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
Molly adjusted the strap of her utility belt, groaning under the weight of three spools of braided nylon, a grappling hook fashioned from a rusted coat hanger duct-taped to a ceramic dinner fork, and a holstered device resembling a hair dryer wired to a calculator. She crouched low behind a topiary hedge sculpted into an aggressive-looking swan, her eye pressed against the single lens of a periscope constructed from cardboard tubes, two mirrors, and superglue. Through the distorted view, the Henderson residence loomed like a fortress of joy. Curtains depicting cartoon dinosaurs danced violently in the draft, silhouetting the frantic movement of adults clutching plates of finger sandwiches. The air hummed with the distinct vibration of excessive sugar consumption. "Perimeter holds," Molly rasped, peering through the magnified lens. "Three minutes until the buffet line forms. Calibrate the floss winch. Hit the jam jar switch. We strike when the candles die." Bill swallowed hard, his welding goggles fogging instantly as he clutched a reinforced canvas duffel bag. The fabric strained taut over the angular outline of remote-controlled cars, his knuckles white beneath latex gloves stained with grease and glitter. He shifted his weight, causing a dry leaf to snap with the volume of a gunshot, and froze before struggling to maintain balance. "On it," Bill hissed. "But what about Mr. Whiskers? The dog knows karate. The balloon swords look lethal. They have pointy ends! He'll bite our ankles!" Molly didn't look up from her knot-tying, her fingers moving rapidly. She tested the tension on the nylon line threaded through the gutter, giving it a sharp tug that vibrated up her arm like a plucked cello string. Satisfied, she pointed a gloved finger toward the skylight where a cluster of helium balloons hovered near the chimney flue. "Mr. Whiskers is sedated by Tuesday's vet visit," Molly muttered, unspooling a roll of micro-fibers. "Besides, we have the yarn grid. Dance the Macarena, don't touch the red strands, and we're golden. If you tangle, you fall. One loose thread and we're stuffed. Do you want to be stuffed? Like a piƱata?" She nudged Bill's shin with her boot, handing him the coiled rope weighted with a marble. The wind shifted, carrying the faint aroma of frosting wafting from the open kitchen window. Molly checked the trajectory, calculating wind resistance and the bounce factor of their descent into the playroom carpet, her mind racing through variables. She raised the grappling hook, aiming the ceramic fork toward the gutter, and waited for the shadow of the cake server to dip toward the table, her heart hammering a war drum rhythm against her ribs as she prepared to launch their fate into the lion's den of unwrapped plastic dreams.
created_at
"2026-09-02T15:07:20.592Z"
9/2/2026, 3:07:20 PM
model
"qwen/qwen3.7-flash-20260727"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
373
moderation_latency
(null)
generation_time
50541
tokens_prompt
75
tokens_completion
7292
native_tokens_prompt
76
native_tokens_completion
7417
native_tokens_completion_images
(null)
native_tokens_reasoning
6800
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
0.00096649
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788361640-QQmh2GR8404B2lPr66Zk"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788361640-Ia6CZiRttKcLC4CzxwM0"
upstream_id
"chatcmpl-78dc3ebb-57b4-9a2d-97fe-c140d9360eff"
provider_responses
0
endpoint_id
"5a9a0ee6-874a-4d03-92be-e5094c0df4e7"
id
"chatcmpl-78dc3ebb-57b4-9a2d-97fe-c140d9360eff"
is_byok
false
latency
373
model_permaslug
"qwen/qwen3.7-flash-20260727"
provider_name
"Alibaba"
status
200
total_cost
0.00096649
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
0.00%
Matches word count
n/a
neededClean
false
words
452
0.99%
Dialogue to Total Word Ratio
Ratio: 20.39%, Deviation: 9.61%
neededClean
false
wordsTotal
456
wordsDialogue
93
0.4961%