NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HHQNMG4Y6JRA7D1XTDV0YB
Run Details
Model:
Qwen 3.8 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
2m 19s
Cost:
$0.008243
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.95%
Dialogue to Total Word Ratio
(details)
99.9751%
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 crouched behind a hedge and studied the birthday house. "The target is twelve presents under a bouncy castle." Bill adjusted his tiny goggles. "Twelve? That's barely a fortune." "Focus," Molly said. "One gift says wooden unicorn. Another says robot. That is basically two cars." Bill wrote on a napkin. "I'm confessing professionally. Suspected dogs." "We are not dogs," Molly hissed. "We are, according to notes, fierce and loyal to loot." Molly pointed at the garden. "Parents distracted by cake. Clown arguing. We enter after piƱata." Bill nodded. "What's the alarm? A dog? A laser?" "A toddler named Kevin," Molly said. "Kevin cries at moved gifts." "Kevin sounds professional." "Kevin is our only security." Bill rubbed his chin. "Distract him with balloons, juice, or an adult." "First rule: never trust a bounce house." "Second rule: leave a pony if you take a unicorn." "We are not leaving a pony," Molly snapped. "Strong thighs." Molly glared. "Stealing all the presents." "All? Evil. Hats too?" "Hats are camouflage." Bill checked his watch. "What if presents contain books?" Molly paused. "Books are for lairs. Check cards." "We cannot steal cards." "Then how know candy?" "We don't. Thrill." Bill drew a plan. "Molly scales hedge. Bill asks bathroom. We grab loot." "Can you ask?" "I have urgent face." "Urgent, not suspicious." "Urgent suspicious since 1997." Molly sighed. "Inside, bag presents quietly." "Robot says thanks if kicked." "Do not kick." "Roll it like bowling ball." "This is disaster with wrapping paper." Bill held up candy-striped gloves. "Matching gloves." Molly looked. "Wrong gloves." "Birthday gloves." "Party Pooper." "Code name." Molly took gloves. "Seven. Signal at cupcakes." "What signal?" "Spit cherry." "No." "Like ruby. Poetry." Molly stared. "We steal, escape bouncy castle, no notices." "Kevin?" "Leave marked Christmas present." "Cruelest plan ever." "Married me?" "Proposing." "With unicorn?" "Unicorn and heart." Bill stood. "For loot." Molly adjusted her sparkly watch. The party lights made the hedge glow like a disco cake. She imagined presents sliding into bags, parents blaming cake fog, and the clown surrendering his nose to fate. Bill stood with knees together, looking like a serious garden gnome. His shoes squeaked twice, and Molly shot him a glare. He froze, holding one finger up. The wind smelled of vanilla and regret. Inside, a song started badly. Molly's heart beat faster. Bill's belt jingled with stolen party noise, tiny bells that should have been left at home. She counted them and added them to the danger list. A moth slapped the garden lamp, proving insects wanted a cut. The bouncy castle shivered behind a tarp. A child shouted for sugar, and Molly felt old. Bill leaned close. "Ready?" Molly stared at him, then at the presents glowing through lace curtains. "Do we take books?" Bill asked. "Only candy books," Molly said. "With pictures." Molly pulled him behind a hedge as frosting light flickered and tiny plastic swords rattled inside the gift pile. She smiled. Bill looked terrified. Party music played badly. A robot laughed inside. They vanished.
created_at
"2026-09-02T17:12:49.118Z"
9/2/2026, 5:12:49 PM
model
"qwen/qwen3.8-flash-20260826"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
5723
moderation_latency
(null)
generation_time
134044
tokens_prompt
75
tokens_completion
14674
native_tokens_prompt
127
native_tokens_completion
17497
native_tokens_completion_images
(null)
native_tokens_reasoning
16692
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
0.00824264
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788369169-AWYb0Nwq1S4HhXK0G0W3"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788369169-4YZfgYks2YtNNddXW8B0"
upstream_id
"chatcmpl-1e76bf41-4ee7-9b46-9a4d-aa0218f2eebe"
provider_responses
0
endpoint_id
"84b1e4a7-5aed-4464-818f-1994f0b4ee18"
id
"chatcmpl-1e76bf41-4ee7-9b46-9a4d-aa0218f2eebe"
is_byok
false
latency
5722
model_permaslug
"qwen/qwen3.8-flash-20260826"
provider_name
"Alibaba"
status
200
total_cost
0.00824264
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
100.00%
Matches word count
n/a
neededClean
false
words
500
99.95%
Dialogue to Total Word Ratio
Ratio: 49.90%, Deviation: 0.10%
neededClean
false
wordsTotal
501
wordsDialogue
250
99.9751%