NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-27-01KQ72C23Z5CBDEKEGCPPG2FGR
Run Details
Model:
Qwen 3.6 35B
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
2m 5s
Cost:
$0.013644
Provider:
AtlasCloud
Evaluation summary
Score
Evaluator
Details Link
99.92%
Matches word count
(details)
91.08%
Dialogue to Total Word Ratio
(details)
95.4982%
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
Molly adjusted her tiara, which pulsed with the low hum of a high-frequency jammer disguised as cheap plastic. She tapped the screen of her smart-glove, checking the thermal feed from the blow-up castle. Bill stood by the snack table, sweating profusely through his 'Dino-Squad' onesie. He clutched a duffel bag woven from stolen party streamers, his knuckles white. He glanced toward a cluster of toddlers waddling like penguins across the linoleum, eyeing a three-year-old reaching for his tail. The room buzzed. Chaos filled the air. Molly hissed, her voice cutting through the roar of screaming children. "Bill, stop fidgeting. Forty seconds to cake. Gluten alerts are a no-go." She rolled her eyes and reached out to steady Bill's trembling shoulder. "He's three. He touches everything. Stay by the piƱata. Don't engage." Bill shrank behind a tower of juice boxes, whispering about bad omens and sticks in eyes. Molly sighed, checking her watch. The second hand spun backward. "Bad omens won't fill the duffel bag, Bill. Remember the mission parameters. We want the high-value loot. The LEGO sets, the video games, and the drone. Ignore the glitter bombs, the socks, and the educational books. We are professionals, not a discount craft store." Bill muttered something about wool blends trending in the underworld. Molly checked her watch again. "Not today. Today is about volume and speed. How's the suction cup packed? We can't afford a mechanical failure when the parents start clapping and the teachers intervene." Bill patted the industrial vacuum hidden inside his onesie. "Packing. I've got the laugh track modification ready. Once the birthday boy opens the master gift, we trigger the noise and sweep the tree. The acoustic dampeners are set to mute the whine." Molly eyed the magician who was currently pulling a handkerchief from his ear. "And the distraction protocol?" "I've replaced the dove with a rubber chicken rigged to squeak at eighty decibels. It's primed in the top hat. Just give the signal when the candle blows." Molly turned to face him, her expression softening just a fraction. She adjusted Bill's unicorn horn, which was slightly askew. "Bill, we're going home wealthy. Unopened Switch." Bill swallowed hard, his Adam's apple bobbing. "I feel uneasy, Moll. My dignity erodes faster than my self-esteem after that tutu incident." Molly grabbed his arm as the first guest of honor approached the cake. "Security's low. Loot is pristine. Check bags." Bill forced a grin that revealed a missing molar and a piece of spinach. It looked less like joy and more like indigestion. "Too much candy. One jelly bean and I'll hum the alphabet song. Focus, Molly. Smile wide. Hide the vacuum hose behind your back and pray." Molly widened his mouth with two fingers until the expression looked manic. "Right. Code word is 'Nap time.' Got it. Let's go steal some birthday magic. And Bill? Listen closely. No scratches. The drone stays sealed. Or I dock your cut and steal your lunch money." "Wouldn't dream of it, partner. Let's raid this shindig."
created_at
"2026-04-27T08:54:52.295Z"
4/27/2026, 8:54:52 AM
model
"qwen/qwen3.6-35b-a3b-20260415"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
825
moderation_latency
(null)
generation_time
124536
tokens_prompt
77
tokens_completion
12474
native_tokens_prompt
76
native_tokens_completion
14122
native_tokens_completion_images
(null)
native_tokens_reasoning
13419
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"
usage
0.013643511
router
(null)
provider_responses
0
endpoint_id
"179c6985-3a1c-48f1-a009-9f3e3e402cdd"
id
"581e6a67384147c7813aece4f9f99b51"
is_byok
false
latency
825
model_permaslug
"qwen/qwen3.6-35b-a3b-20260415"
provider_name
"AtlasCloud"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777280092-TsNZUW7v8MVfhneNoGJD"
session_id
(null)
api_type
"completions"
id
"gen-1777280092-8JmVwvCIIH6gl1A2IjA4"
upstream_id
"581e6a67384147c7813aece4f9f99b51"
total_cost
0.013643511
cache_discount
(null)
upstream_inference_cost
0
provider_name
"AtlasCloud"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
99.92%
Matches word count
n/a
neededClean
false
words
503
91.08%
Dialogue to Total Word Ratio
Ratio: 51.37%, Deviation: 1.37%
neededClean
false
wordsTotal
512
wordsDialogue
263
95.4982%