NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36R520P40670CNSE8JM3GHX
Run Details
Model:
GPT-6 Luna (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
41.8s
Cost:
$0.001742
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.49%
Matches word count
(details)
89.10%
Dialogue to Total Word Ratio
(details)
44.7980%
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 studied the community hall’s cheerful banner, which announced THE BIG BIRTHDAY BASH in letters large enough to alarm passing pilots. Through the open curtains, a mountain of presents glittered beneath paper stars. Beside it stood a cardboard clown, smiling with the confidence of someone who had never been asked to help move furniture. Bill adjusted his bow tie. “Remember,” Molly whispered, “we take the presents, not the cake. The cake has witnesses.” “Last week you stole a traffic cone and apologized to it.” “It looked lonely.” “I can distract the clown.” “It’s cardboard, Bill.” “Then it has no feelings.” “Exactly what you said about the cone.” Their disguises were two borrowed tuxedo jackets, a pair of enormous foam ears, and the sort of confidence usually seen in men who had mistaken a revolving door for a challenge. Molly carried a violin case. It contained no violin, only sandwiches, because she believed every heist benefited from sensible catering. Bill carried a balloon shaped like a giraffe, which kept pointing toward the presents. “Phase one: we enter as entertainers,” Bill said. “You can’t juggle.” “I can drop three objects at once.” “That’s gravity.” “Phase two: gifts into the hamper.” “This is a party, not a hotel.” “Then we carry them.” “Like burglars?” “We’ll wear party hats.” “Still burglars.” At nine, they walked through the already-open hall door behind a parade of children chasing bubbles. Molly’s disguise passed inspection because nobody inspected it; Bill’s balloon snagged the banner and announced their arrival with a long rubber squeal. The cardboard clown toppled onto him. A little girl applauded. Pinned beneath its painted grin, Bill waved as if this were part of the act. Molly began scooping presents, then noticed each tag bore a child’s name and a carefully drawn heart. “Maybe,” Bill whispered, “we take the presents, sell them, and retire.” “Retire where?” “Somewhere with no children.” “Everywhere has children.” “Then we’ll need a bigger hamper.” At the top of the pile lay a crooked package tagged FOR THE CHILDREN’S WARD, decorated with a crayon picture of a smiling moon. Molly paused. Beneath it, the birthday girl was distributing cupcakes to the other children, including one to Bill, who accepted it with the solemnity of a royal treaty. His eyes met Molly’s. The balloon giraffe drooped between them, looking ashamed of its career. “New plan,” Molly murmured. “We steal the cupcakes.” “Too late,” Bill said. “I’ve been given one.” Molly placed every present back in the pile, though she arranged them more neatly than she had found them. Bill surrendered his cupcake to the birthday girl’s little brother, who awarded him a plastic medal for excellent clown recovery. They left with empty hands, two foam ears, and the cardboard clown balanced upright. “Next time,” Molly said, “we rob a piñata.” Bill nodded. “One with a lawyer.”
created_at
"2026-09-23T09:04:38.948Z"
9/23/2026, 9:04:38 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
6246
moderation_latency
(null)
generation_time
41577
tokens_prompt
75
tokens_completion
1478
native_tokens_prompt
66
native_tokens_completion
3471
native_tokens_completion_images
0
native_tokens_reasoning
2833
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
0.0017421
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790154278-no5KvRsf0FsRPAJ8jvkX"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790154278-LCgBQfW4hjATlXd7yG0q"
upstream_id
"resp_0bba0f05608e4083016ab396272a2087d19098602e3c5a6ada"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_0bba0f05608e4083016ab396272a2087d19098602e3c5a6ada"
is_byok
false
latency
2463
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0017421
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
0.49%
Matches word count
n/a
neededClean
false
words
473
89.10%
Dialogue to Total Word Ratio
Ratio: 28.48%, Deviation: 1.52%
neededClean
false
wordsTotal
474
wordsDialogue
135
44.7980%