NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1M26163JW7BSA059Y3J1HPG
Run Details
Model:
Qwen 3.8 27B (Reasoning, XHigh)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
13m 12s
Cost:
$0.058128
Provider:
Reka
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 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
"Tonight we raid the smallest heist target in the county and every single present will vanish before the cake song ends with zero loud noise." Molly said, smoothing her glitter wig, while Bill taped a paper crown to his forehead and spread the party map over the bright van hood. "The clown is security, but he smells frosting and forgets his own birthday, so we only need one distraction with a silly trumpet and cake." Bill said, tapping the map where a balloon cluster guarded the prize table near the punch bowl and the clown with his giant red nose. "Inside the party, we use the hallways like river tunnels, pass the napkin bins, and fill the bag with shiny boxes quietly before the lights." Molly said, placing three fake tickets in her pocket and two tiny flashlights in the oversized clown hat for maximum stealth and quiet style now. "My walk says I belong here, but my eyes say I am stealing the biggest box hidden under the table with a bow and tape." Bill said, rehearsing his innocent walk past the juice table while Molly lifted a toy car with one gloved finger as if sleeping very gently. "If a child asks us for help, we point at the cake and smile like friendly professionals then keep walking away toward the van door." Molly said, pausing by a window where a birthday banner stretched across the wall like cheerful crime scene tape over the snack tray nearby now. "Remember, the presents are heavy, but our van is light, and the alarm is just a squeaky toy with a big red button on top." Bill said, checking his watch, and muttered about cake frosting sticking to their old hidden loot compartment which they had lined with foil papers today. "I go first, because I have the smallest shoes and the best balance for tiptoeing over glitter and broken confetti pieces scattered on the floor." Molly said, pulling a tiny ladder from the trunk and placing it beside the sliding door with care like a museum expert in the dark. "Even the weather is helping us, because hats are useless evidence and fire hydrants never talk to police unless sprayed with water all night long." Bill said, holding the bag open while a party hat blew across the pavement and landed on a fire hydrant with a soft pop nearby. "Now we act like lost aunts who missed the entrance and want to find the best present table before the clown appears with his whistle." Molly said, whispering that the party guests were arriving, while their old engine hummed a nervous little tune under their breath in the cold air. "For the first time in our criminal careers, this job smells like sprinkles, plastic dinosaurs, and total victory plus maybe a cake for later tonight." Bill said, clinking his paper cup, wishing luck, and stepping toward the bright doorway of the party while holding the stolen map now with care.
created_at
"2026-09-03T16:38:40.642Z"
9/3/2026, 4:38:40 PM
model
"qwen/qwen3.8-27b-20260814"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
313
moderation_latency
(null)
generation_time
792157
tokens_prompt
75
tokens_completion
20154
native_tokens_prompt
118
native_tokens_completion
22784
native_tokens_completion_images
(null)
native_tokens_reasoning
19446
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
0.05812752
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788453520-LqXF3W9a0jkMYEU24t6a"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788453520-gvd6dpg2lwTuzszEDUlt"
upstream_id
"chatcmpl-e44a917d-fcbc-41a2-b00a-eef469c9126f"
provider_responses
0
endpoint_id
"15086968-df2f-4d69-86bb-c74c31644bb6"
id
"chatcmpl-e44a917d-fcbc-41a2-b00a-eef469c9126f"
is_byok
false
latency
311
model_permaslug
"qwen/qwen3.8-27b-20260814"
provider_name
"Reka"
status
200
total_cost
0.05812752
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Reka"
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: 50.00%, Deviation: 0.00%
neededClean
false
wordsTotal
500
wordsDialogue
250
100.0000%