NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36P1WR3V2VFX8PSBFEA8T9Q
Run Details
Model:
GPT-6 Luna (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
18.8s
Cost:
$0.000617
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
90.48%
Matches word count
(details)
98.91%
Dialogue to Total Word Ratio
(details)
94.6956%
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 and Bill crouched behind a hedge outside the community hall, watching a banner flap above the door: HAPPY BIRTHDAY, LEO! Beneath it, children charged in circles while adults carried trays of cupcakes. A mountain of presents occupied the stage. Bill consulted a napkin covered in arrows, none of which pointed in the same direction. “Once we’re inside, we take the presents and leave,” he whispered. “Excellent,” Molly said. “A plan with the elegance of a shopping trolley.” “I’m being practical.” “You’ve drawn a trapdoor.” “That’s the cake table.” Molly peered through the window. A clown was teaching the children to juggle oranges. One orange bounced off a lampshade, and the children applauded as if gravity had performed a magic trick. “How do we get past the adults?” Bill asked. “We could pretend we’re magicians.” “You can’t do magic.” “I can make a wallet disappear.” “Not from someone else’s pocket, Molly.” “That’s a matter of perspective.” Their disguises were a pair of borrowed party hats and Bill’s old tuxedo jacket, which had a streak of mustard down one sleeve. They entered with a cluster of arriving relatives, carrying a gift bag between them. Inside it was a brick, included for what Bill called “authentic heft.” At the door, a woman handed them name tags. Molly wrote “Auntie” on hers. Bill wrote “The Magician,” then crossed it out and wrote “Uncle.” A little boy examined him. “Which one?” “Your uncle,” Bill said. “From where?” Bill glanced at Molly. “The north.” The boy nodded solemnly and ran off, apparently satisfied that the north contained uncles. They edged toward the stage. Bill lifted the first present, a box wrapped in silver paper. A tiny bell jingled inside. “Careful,” Molly murmured. “I am careful.” “You’re holding it upside down.” “I’m carefully holding it upside down.” A child appeared beside them, wearing a cardboard crown. “Are you taking Leo’s presents?” Molly froze. Bill lowered the box. “We’re checking for dangerous toys,” Molly said. “Is that true?” “It’s almost true,” Bill whispered. The child considered this, then pointed at the brick-filled gift bag. “That one looks dangerous.” Before either thief could respond, the clown bounded over and plucked the bag from Bill’s hand. “A surprise for the birthday boy!” he announced, placing it on the gift pile. Bill watched their brick join the presents. “We’ve donated our getaway weight.” “Maybe it’ll be useful,” Molly said. “He can build a shed.” The children began singing. Leo stood before the presents, grinning, while his parents filmed. Molly looked at the wrapped boxes, then at the boy’s bright, expectant face. “New plan,” she whispered. “Steal the cake?” “Help him open everything.” Bill sighed. “That plan has terrible profit margins.” They joined the crowd. Molly handed Leo the silver box. Bill passed out cupcakes. When Leo unwrapped the brick, everyone stared. Bill cleared his throat. “It’s a very sturdy toy.” Leo hugged it. “I’m building a castle!” Molly smiled. Bill eyed the exit, then took another cupcake. For once, the best thing they’d stolen was the chance to leave empty-handed.
created_at
"2026-09-23T08:27:58.089Z"
9/23/2026, 8:27:58 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
6969
moderation_latency
(null)
generation_time
18736
tokens_prompt
75
tokens_completion
1135
native_tokens_prompt
66
native_tokens_completion
1221
native_tokens_completion_images
0
native_tokens_reasoning
512
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
"completed"
service_tier
"default"
usage
0.0006171
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790152078-d7blrmM2pKQBRuMQQyBD"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790152078-zDJWnddtnKIC2p8Zg1ce"
upstream_id
"resp_0290cfaa3e15f646016ab38d8e305887d1bbc8b0416b6f28c8"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_0290cfaa3e15f646016ab38d8e305887d1bbc8b0416b6f28c8"
is_byok
false
latency
737
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0006171
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
90.48%
Matches word count
n/a
neededClean
false
words
510
98.91%
Dialogue to Total Word Ratio
Ratio: 30.47%, Deviation: 0.47%
neededClean
false
wordsTotal
512
wordsDialogue
156
94.6956%