NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HK2FBAVJ4SJ52QDME26ABZ
Run Details
Model:
Qwen 3.8 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
2m 19s
Cost:
$0.007595
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
99.92%
Matches word count
(details)
99.49%
Dialogue to Total Word Ratio
(details)
99.7021%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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 van smelled like cheese and regret. Molly spread a crayon drawing across the dashboard. The target was a suburban backyard. The gate was painted blue and locked only by habit. Two folding chairs guarded the cake table like plastic soldiers facing a war they would lose. Molly had studied every birthday party on the block. She knew the parents bought identical drinks, identical napkins, and identical mistakes. Bill had studied only the bakery window, but he was learning. Their wagon sat in the grass, already waiting with great appetite. "Is this the party plan?" Bill squinted at a stick figure vomiting cake. "It looks like a hostage negotiation with frosting." "It’s the layout," Molly said. "Fourteen bouncy houses, one ball pit, and a piñata shaped like the mayor." "God bless democracy," Bill muttered. "Where are the presents?" "Under the inflatable dragon, left side, near the cotton-candy cart." "Excellent. We arrive as clowns?" "Mime duo. Silent, unsettling, legally untrackable." Bill clapped once. "I can be silent." He leaned in. "What’s our cover?" "Entertainment. I do face painting. You handle distractions." "I handle distractions?" "You cry loudly." "Ah. A specialty." Molly tapped the diagram. "When the parents are distracted, we load the presents into the wagon." "A wagon with a unicorn sticker and the words 'Emergency' in marker." Bill considered. "Can we steal the unicorn too?" "Focus." "Right. Timeline?" "Party starts two. Presents opened three. We strike at two fifty, because that’s when the magician fails his first trick and everyone checks their phones." "The magician is our inside man?" "He’s my brother." Bill whistled. "Is he reliable?" "He’s a magician. He lies for a living." "Then he’s perfect." Molly rolled the map. "We need gear." "No, glitter." Bill gasped. "Glitter is diabolical." "It gets everywhere, ruins rugs, and parents flee the house." "Then we take the presents while they vacuum?" "You’re a monster." "I’m efficient." She checked her watch. "You’re carrying the decoy cake." "It’s fake. Foam. But if the clown walks in with cake, people watch him." "I’m wearing foam?" "No, you’re carrying it near your chest." "And if I fall?" "Then we lose a great man and one very convincing lemon." Bill stood, nearly hitting his head. "I’m going to need gloves." "Latex. We don’t want fingerprints on the toys." "What if a toddler sees us?" "Offer a balloon." "What if he wants a dinosaur?" "Give him a balloon dinosaur." She tossed him a mask. "One more thing. No stealing the piñata." Bill caught it, offended. "I was going to ask." "You have a weakness for candy." "I have a weakness for opportunity." "That’s worse." Molly opened the van door. "Let’s go. Dragon under, wagon ready, tears everywhere." "Shall we call it Operation Gift Snatch?" "We shall not." "What about Operation Birthday Bandit?" "We are professionals." Bill climbed out. "Then what’s the name?" Molly smiled. "Operation Free Cookies." "There are no cookies." "There are always cookies at children’s parties." "Then our getaway will be delicious."
created_at
"2026-09-02T17:36:06.514Z"
9/2/2026, 5:36:06 PM
model
"qwen/qwen3.8-flash-20260826"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
733
moderation_latency
(null)
generation_time
138486
tokens_prompt
75
tokens_completion
13301
native_tokens_prompt
127
native_tokens_completion
16120
native_tokens_completion_images
(null)
native_tokens_reasoning
15361
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
0.00759545
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788370566-ey2lCfdcKogSix6FadNS"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788370566-HFDLkiEKUvqKwOtTUm7I"
upstream_id
"chatcmpl-d28bddc2-5ae6-9249-bd95-799dec4552e5"
provider_responses
0
endpoint_id
"84b1e4a7-5aed-4464-818f-1994f0b4ee18"
id
"chatcmpl-d28bddc2-5ae6-9249-bd95-799dec4552e5"
is_byok
false
latency
733
model_permaslug
"qwen/qwen3.8-flash-20260826"
provider_name
"Alibaba"
status
200
total_cost
0.00759545
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
99.92%
Matches word count
n/a
neededClean
false
words
497
99.49%
Dialogue to Total Word Ratio
Ratio: 69.68%, Deviation: 0.32%
neededClean
false
wordsTotal
498
wordsDialogue
347
99.7021%