NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36NN5PC3CH1R8HZMA2B3NYW
Run Details
Model:
GPT-6 Luna (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
19.0s
Cost:
$0.000616
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
75.16%
Matches word count
(details)
99.95%
Dialogue to Total Word Ratio
(details)
87.5542%
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
At nine o’clock, Molly and Bill crouched behind a hedge outside the community hall, watching a banner flap in the wind: HAPPY BIRTHDAY, LEO! Beneath it sat a window display of presents, piled high as a brightly wrapped mountain. “Tonight,” whispered Molly, “we take every present. All of them. We leave no ribbon unplucked.” “Shouldn’t a heist have a more dignified target?” Bill asked. His fake mustache drooped into his tea. “Dignity is for people with steady jobs.” Bill peered at the hall. Children shrieked inside, a sound like a flock of delighted seagulls discovering a trampoline. “I’m still unclear on the escape plan.” “We walk out carrying presents.” “That’s not a plan. That’s carrying presents.” “It’s a plan with excellent posture.” Molly had selected the party because she considered children’s celebrations “lightly guarded.” She had not anticipated the security arrangements: three vigilant parents, one enormous dog wearing a paper crown, and a six-year-old stationed by the door with a clipboard. He checked names, gifts, and, inexplicably, shoe size. Bill adjusted his mustache. It fell off and landed in a puddle. “Abort,” he said. “Don’t be ridiculous. We’re already disguised.” “As what?” “Adults.” “That’s our normal disguise.” Molly produced two party hats from her coat. One said BIRTHDAY BOY; the other, in glittering letters, said I’M HERE FOR THE CAKE. She put on the cake hat and handed Bill the other. “I refuse to impersonate the birthday boy,” he said. “You’re not impersonating him. You’re celebrating his leadership.” The clipboard child spotted them. “Are you on the list?” Molly smiled with the confidence of someone who had once returned a library book on time. “We’re the entertainment.” The boy studied them. “What do you do?” Bill, under pressure, performed a card trick. He produced the three of diamonds from behind Molly’s ear, then discovered it was actually a wet leaf. “We’re ventriloquists,” Molly announced. “Then why are you talking?” asked the boy. Bill opened his mouth. Molly quickly added, “He’s the dummy.” A parent appeared in the doorway, holding a tray of cupcakes. “You two must be the clowns.” “We are?” Bill murmured. “Apparently,” Molly murmured back. “Try to look funny.” Bill slipped on the wet leaf. The children applauded. The dog barked. The parent handed them each a cupcake, and the clipboard boy waved them inside. Molly glanced toward the presents. Her original plan shimmered before her like a diamond—or perhaps a very shiny felony. Then Leo, the birthday boy, offered her a balloon sword. “Can you make a balloon dinosaur?” he asked. Molly looked at Bill. Bill looked at the cupcake. The cupcake had sprinkles. “Certainly,” she said. Twenty minutes later, they were surrounded by children, balloon animals, and an alarming number of sticky fingerprints. The presents remained untouched. As they left, Bill carried a paper plate stacked with cake. “So,” he said, “we stole nothing.” “We stole their hearts.” “I’d rather have stolen another cupcake.” Molly considered this. “Next time, we plan better.” From inside, Leo shouted, “The clowns forgot their mustache!” Bill sighed. “Next time, we wear two.”
created_at
"2026-09-23T08:21:01.266Z"
9/23/2026, 8:21:01 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
4408
moderation_latency
(null)
generation_time
18951
tokens_prompt
75
tokens_completion
1259
native_tokens_prompt
66
native_tokens_completion
1218
native_tokens_completion_images
0
native_tokens_reasoning
488
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
0.0006156
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790151661-vJ7C02HopKCnxXKsSVcP"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790151661-qBfpaRm2E6v0gUEaYtS8"
upstream_id
"resp_08929b4587490618016ab38bed5a6c87d1bccf752532dac790"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_08929b4587490618016ab38bed5a6c87d1bccf752532dac790"
is_byok
false
latency
350
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0006156
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
75.16%
Matches word count
n/a
neededClean
false
words
513
99.95%
Dialogue to Total Word Ratio
Ratio: 30.10%, Deviation: 0.10%
neededClean
false
wordsTotal
515
wordsDialogue
155
87.5542%