NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1M1D836Q2CNZ48E431AA875
Run Details
Model:
Qwen 3.8 27B (Reasoning, XHigh)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
2m 57s
Cost:
$0.026428
Provider:
Phala
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 30% 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 beside the rental van while Bill adjusted his party hat to look less like a man and more like a disappointed uncle. "Bill, if we steal every present, the party becomes a crime scene made of glitter, balloons, and small, accusing eyebrows." Across the lawn, the Hendersons were holding a party for eight children, six grandparents, and one extremely confident puppy. Streamers twisted in the breeze like thin red flags of surrender. A banner read Happy Birthday Leo in letters that had survived exactly three hours of wind. The presents lay in the gazebo, wrapped in shiny paper, tied with ribbons, and arranged with the confidence of furniture that expected to be judged. "We take the toys, Bill, not the cupcakes, though I suspect the cupcakes are the real treasure, and I am not above a bite." Molly opened the trunk and revealed a kit of tools that belonged in a mystery novel written by a man who trusted nothing, especially birthday parties. She lifted a lockpick, a silk scarf, a tiny mirror, and a bag of plastic coins that looked suspiciously like real change except for the fact that every coin said LOL on one side. "The parents are in the garden judging each other's hats, the kids are in the hall chasing a plastic dinosaur, and the presents are stacked behind a curtain like a very colorful vault." Bill checked his watch, then checked the hedge, then checked the hedge again, as if repetition could make the operation look less ridiculous. He smoothed his cardigan, pulled a birthday noose from his pocket, and folded it with the tenderness of a surgeon preparing for a small, sweet tragedy. "We enter as cleanup volunteers, you carry the big box, I carry the tiny box, and our signal is three polite coughs from near the gift table." The gazebo door had one lock and a brass knob shaped like a smiling starfish. Molly nodded at it with the respect one gives to a minor obstacle. Bill nodded back, because in heist work, nodding is often the loudest sound a thief makes before getting caught. "If a child asks questions, you become the tooth fairy, and if a parent asks questions, we become the tooth fairy with a van." They slipped through the hedge, over the toy box, and into the gazebo. Each wrapped package stood like a small monument to childhood greed. Bill placed the big box on the floor, and the tiny box chose itself a spot beside a wrapped present with a rocket ship on it. "Good plan, Bill, because I have already hidden the glitter bag, the ribbon scissors, and the emergency donut under the passenger seat." Molly coughed once, twice, three times. Bill understood. He lifted the big box while Molly lifted the tiny box, and together they moved toward the van with the solemn joy of thieves who had not yet considered whether the presents were heavier than fame before the party could end.
created_at
"2026-09-03T16:25:08.462Z"
9/3/2026, 4:25:08 PM
model
"qwen/qwen3.8-27b-20260814"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
366
moderation_latency
(null)
generation_time
176507
tokens_prompt
75
tokens_completion
7029
native_tokens_prompt
186
native_tokens_completion
8792
native_tokens_completion_images
(null)
native_tokens_reasoning
8193
native_tokens_cached
64
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
0.026428
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788452708-LwtUt2XbRIWcszkiEavY"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788452708-O42uCc1gjTUbL4aI3yN0"
upstream_id
"req_41ec4c3abd24fbac09c33a8795f32492"
provider_responses
0
endpoint_id
"cfc4d0ca-df7c-4b5f-9c74-328c41a7cbc3"
id
"req_41ec4c3abd24fbac09c33a8795f32492"
is_byok
false
latency
366
model_permaslug
"qwen/qwen3.8-27b-20260814"
provider_name
"Phala"
status
200
total_cost
0.026428
cache_discount
0.0000224
upstream_inference_cost
0
provider_name
"Phala"
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: 30.00%, Deviation: 0.00%
neededClean
false
wordsTotal
500
wordsDialogue
150
100.0000%