NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HJ7TG684S4B3TQWFQ6E8ZH
Run Details
Model:
Qwen 3.8 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
4m 45s
Cost:
$0.012798
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
Molly flattened a napkin diagram on the table. “Tonight’s mark: the Pemberton birthday party.” Bill adjusted a feathered hat. “Why are we stealing presents?” “Because children are terrible at security.” Molly pointed to the diagram. “They have parents.” “Parents are easily confused by laminated badges.” Bill lifted a plastic scanner. “And glitter.” “Exactly. Glitter is chaos in a cute outfit. Our target is the table under the balloon arch.” Molly handed him silly gloves. Bill dropped one glove. “How many presents?” “Sixty, wrapped, fragile, and emotionally explosive.” Molly taped a fake badge to his chest. “Do we take them all?” “We take every ribboned box.” Bill tried to look official. “What do we leave behind?” “Decoys full of socks, coupons, and disappointment.” A child’s laugh echoed from the doorway. “That’s evil.” “That’s brand management.” Molly leaned closer. “I like it.” “The bouncy castle is our exit route.” Bill pulled a rubber chicken from his pocket. “Do we bounce?” Molly ignored it. “We bounce only when the parents start looking.” Bill practiced dropping a tray. “Can I pretend to be a clown?” “No. You are my charming gift inspector.” The cart squeaked as Molly tested its wheels. “I can inspect aggressively.” Bill held the decoy box open. “First, you drop a tray near the cake.” Molly checked a timer shaped like a lollipop. “Then chaos blooms like fondant flowers.” Bill nodded, then frowned. “And while everyone panics, we wheel out the cart.” Molly tucked foil around a remote. “That’s surprisingly simple.” “Simple is expensive. We are broke.” The fake gifts glittered like cheap treasure. “I didn’t think about being broke.” Bill grabbed the cart. “Focus, Bill. The birthday child is Theodore.” Molly opened the door. “Theodore sounds innocent.” Her coat pockets bulged with wrapping paper, ribbon, and guilt. “Theodore is three. He cannot tell a lawyer from a lamp.” Bill straightened the feather in his hat. “That is a terrible thing to know.” Molly checked the frosting diagram again. “Use it at parties.” They moved together like professional idiots. Molly nodded once. “What if Theodore sees the cart?” “Then you say it is a surprise inspection.” “A surprise inspection of birthday presents?” “Of course. Presents must meet legal standards.” “Whose legal standards?” “Ours, obviously.” “You are frighteningly organized.” “Stealing candy from children requires a spreadsheet.” “Do you have a spreadsheet?” “I have a napkin with frosting.” “That is not a spreadsheet.” “It is a spreadsheet with emotional flavor.” “Fine. What is my disguise?” “You wear the tuxedo and the mustard.” “That is not a disguise.” “It is a story.” “What story?” “The elegant waiter who has seen too much dessert.” “I like dessert too much.” “Good. You also have the rubber chicken.” “Is it a distraction?” “It is my emotional support comedy.” “What happens if security comes?” “We smile, wave, and blame the cake.” “I can blame a cake.” “I know.” “Are we ready?” “No. But ready is for rich people.” “Then I’m stealing presents like a poor professional.” “Perfect.” “Go.”
created_at
"2026-09-02T17:21:39.588Z"
9/2/2026, 5:21:39 PM
model
"qwen/qwen3.8-flash-20260826"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
713
moderation_latency
(null)
generation_time
278110
tokens_prompt
75
tokens_completion
21832
native_tokens_prompt
127
native_tokens_completion
27190
native_tokens_completion_images
(null)
native_tokens_reasoning
26410
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
0.01279835
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788369699-3jtOcreeG1RfyO9wDRRf"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788369699-52kq3vAWLJNfqrkhepDd"
upstream_id
"chatcmpl-35b07e26-b692-9c17-9ebe-09b52ecd9e87"
provider_responses
0
endpoint_id
"84b1e4a7-5aed-4464-818f-1994f0b4ee18"
id
"chatcmpl-35b07e26-b692-9c17-9ebe-09b52ecd9e87"
is_byok
false
latency
713
model_permaslug
"qwen/qwen3.8-flash-20260826"
provider_name
"Alibaba"
status
200
total_cost
0.01279835
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%