NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HH4GWRY8HGQXNKHVVN3W5Z
Run Details
Model:
Qwen 3.8 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
3m 13s
Cost:
$0.011368
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
100.0000%
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 plan is simple," Molly said, tapping a crayon against a folded birthday invitation. Bill studied the backyard, where a banner sagged between two trees. "Simple? You drew a map on a pancake." "It's a stack of pancakes, and this one is the gift table." "And the syrup?" "That's security." Bill lowered his binoculars. "How are we getting past the clown?" "We don't get past him. We get around him." "By going through the bouncy castle?" "By becoming bouncy." Molly pulled a rubber sheet from her coat. "It's a spare inner tube for a castle, a very small child, or a moral decision." Bill poked it. "Is this inflatable?" "Only if we need it to be." "That's not physics." "It's comedy." Molly pointed to a house with a red roof. "Presents will be behind the kitchen door." "Kitchen doors are for cooks." "Cooks are thieves who forgot their calling cards." Bill smiled, then frowned at the cake. "What if the child cries?" "Then the party has improved." "You mean it will have drama." "I mean it will have witnesses." Bill opened a notebook with a jelly stain on page one. "Timing?" "At two, the clown honks, and everyone looks at the banner." "The banner?" "The big one. It says Sir Squinty is Three." "Sir Squinty sounds like a villain." "Sir Squinty is the birthday llama." Bill blinked. "There is a llama?" "A plush one, but I don't trust it." Molly checked her watch. "We enter through the laundry room." "Why laundry rooms?" "They smell like secrets and socks." Bill folded the map pancake into his pocket. "What about the parents?" "Parents are a challenge, a mystery, and a casserole threat." "Can we steal the casserole too?" "If it fits in the sack." "How many presents should we take?" "All of them." Bill counted on his fingers, then stopped at ten. "I don't have enough fingers." "We can count by pairs." "I can count by snacks." "This is a heist, not a picnic." "Every heist is a picnic for someone." Molly held up a velvet sack embroidered with stars. "The sack is for the loot, not the snacks." "What if we leave one present?" "Never." Bill tucked the invitation into his hatband. "I respect the rule." "Also, no squeaky shoes." Bill looked down at his boots. "Why do you always choose the funny shoes?" "Silence is funnier." "You're no fun." "I'm exactly enough fun to keep us from prison." Molly checked the clock on a birthday cake stand. "Two minutes." Bill put on a plastic tiara. "This is for disguise." "You look like a confused librarian." "Confused librarians are invisible." "Do we steal the pinata too?" "Only if it contains toys." "That seems greedy." "Professional." Molly tucked her fake mustache behind a favor. "We go at two." Molly tugged on a tiny clown hat. "We will look harmless." Bill wore the same hat and looked worse. Bill adjusted his hat. "We look harmless for people who have never committed a crime."
created_at
"2026-09-02T17:02:16.48Z"
9/2/2026, 5:02:16 PM
model
"qwen/qwen3.8-flash-20260826"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
614
moderation_latency
(null)
generation_time
192597
tokens_prompt
75
tokens_completion
17412
native_tokens_prompt
127
native_tokens_completion
24146
native_tokens_completion_images
(null)
native_tokens_reasoning
23393
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
0.01136767
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788368536-8Id6Ane8zyNcJUazAy9z"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788368536-x6PnafnXCGDsppzp0n9P"
upstream_id
"chatcmpl-ab2ae66c-f884-9879-bf08-aa5a0fada608"
provider_responses
0
endpoint_id
"84b1e4a7-5aed-4464-818f-1994f0b4ee18"
id
"chatcmpl-ab2ae66c-f884-9879-bf08-aa5a0fada608"
is_byok
false
latency
614
model_permaslug
"qwen/qwen3.8-flash-20260826"
provider_name
"Alibaba"
status
200
total_cost
0.01136767
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
100.00%
Dialogue to Total Word Ratio
Ratio: 70.00%, Deviation: 0.00%
neededClean
false
wordsTotal
500
wordsDialogue
350
100.0000%