NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36R1NP7RJM858P136NMPMCV
Run Details
Model:
GPT-6 Luna (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
1m 42s
Cost:
$0.005649
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
99.98%
Matches word count
(details)
99.55%
Dialogue to Total Word Ratio
(details)
99.7684%
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
Molly spread a crumpled map across the van’s dashboard, pinning its corners with a thermos, a stale muffin, and Bill’s dented emergency mustache wax tin. “Bill, presents are inside the community hall, guarded by a clown, two parents, and a balloon shaped like a tax auditor. Excessive for a six-year-old.” Bill leaned over Molly’s sketch of the hall, squinting. It showed a door, a window, and something labeled “possibly cake” in the far back corner. “Molly, we enter through the kitchen. Every party has a kitchen, and every kitchen has a door propped open by somebody carrying frosting at speed.” Molly tapped a blue crayon route that crossed the parking lot, skirted a hedge, and ended at a rectangle she had drawn with suspicious confidence. “That rectangle is the kitchen?” “It was before the map got damp.” “From what?” “Your muffin, apparently, has the structural integrity of soup.” Across the street, the party hall glowed with paper stars. A sign announced NED’S SIXTH BIRTHDAY, though someone had changed the N to an H. “Hed’s birthday. Sounds like a sheep’s party. Bill, adjust your tie; you can blend in. Tell them you sold raffle tickets at a wool convention.” He wore a paper crown from an earlier operation, though the glitter had migrated into his eyebrows, making him look permanently surprised and strangely regal. “Presents are our target. Not cake, not piñata, and definitely not children.” “I hadn’t listed them, Molly; I meant clown as a target, strictly speaking.” Molly produced two empty tote bags. One read LIBRARY BOOKS; the other read I’M WITH THE BAND, despite neither thief knowing an instrument at all. “We lift every wrapped box and leave before anyone notices.” “What if a child notices?” “Then we compliment the wrapping with absolutely sincere admiration, naturally.” Bill tested the van’s side door. It squealed like a haunted kettle, so Molly shut it before the neighborhood cats could assemble outside for judgment. “We need a distraction. I could juggle, Molly, confidently.” “Can you juggle?” “No, but the children will be impressed by my confidence and brave posture.” Molly glanced at the party’s window, where a small face appeared, vanished, then reappeared wearing a paper mustache and a serious expression of quiet judgment. “That child has spotted us.” “Impossible. He’s six.” “Six-year-olds are observant. They once caught me hiding broccoli in my napkin during family dinner last Sunday.” Bill lowered himself behind the dashboard. His crown slipped over his eyes, turning his whispered strategy into a muffled speech addressed to his own lap. “New plan: Molly and I enter as entertainers, perform a trick, and make the presents disappear.” “Bill, that is the original plan with louder applause.” Molly folded the map and tucked it away. From the hall came a chorus singing off-key, followed by a tremendous crash and adult applause somewhere. “The children have started the piñata. Operation Gift Grab is canceled.” “Because of morals?” “Because, Bill, we hate crowds, balloons, and screaming more than bills.”
created_at
"2026-09-23T09:02:48.018Z"
9/23/2026, 9:02:48 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
5048
moderation_latency
(null)
generation_time
101587
tokens_prompt
75
tokens_completion
1792
native_tokens_prompt
66
native_tokens_completion
11284
native_tokens_completion_images
0
native_tokens_reasoning
10609
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
0.0056486
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790154168-iVHeGjo8IkNltM1oiWd6"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790154168-7v7ycLJxIUkOq0qcX51d"
upstream_id
"resp_03cbd4a144138d5a016ab395b8273c87d19c5e94432639ccf5"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_03cbd4a144138d5a016ab395b8273c87d19c5e94432639ccf5"
is_byok
false
latency
1472
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0056486
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
99.98%
Matches word count
n/a
neededClean
false
words
498
99.55%
Dialogue to Total Word Ratio
Ratio: 50.30%, Deviation: 0.30%
neededClean
false
wordsTotal
501
wordsDialogue
252
99.7684%