NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36NX69EWDX9T3NWH7K19JMF
Run Details
Model:
GPT-6 Luna (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
58.6s
Cost:
$0.003258
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.95%
Dialogue to Total Word Ratio
(details)
99.9751%
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 and Bill crouched behind a hedge, watching the community hall glow like a birthday cake with windows, balloons, and absolutely no respectable escape plan. “The presents are piled beside the punch bowl, and nobody suspects two professionals disguised as ordinary adults with unusually large gift bags tonight, over there.” Bill adjusted his paper crown, which read Birthday VIP, as they slipped through the open doors behind a parade of parents and balloon animals inside. “I prefer ‘guest of honor.’ VIP sounds like a person who knows where the bathrooms are and can operate a juice box independently without supervision.” Inside, children zigzagged between folding chairs while a magician produced rabbits, scarves, and the increasingly worried expression of a man whose rabbit had escaped again. “We wait for the cake ceremony. When everyone sings, the noise will cover our footsteps, unless they sing in rounds like trained owls wearing hats.” Bill checked their disguises: Molly wore a cardigan covered in tiny ducks, and he wore a tie shaped like a nervous banana, yesterday’s breakfast mascot. “If anyone asks, we’re cousins from out of town, attending our niece’s party and carrying these bags for charitable reasons today, naturally, with impeccable manners.” A small boy approached, carrying a plastic dinosaur and the grave authority of someone who had already questioned three adults about suspicious snacks twice today. “If anyone calls us present inspectors, we’re gift consultants specializing in ribbon quality and sudden exits when the music stops during important ceremonies especially indoors.” Bill nodded solemnly, then accidentally saluted the dinosaur, which seemed unimpressed but willing to maintain professional confidentiality for a modest cookie and one juice box. “No cookies? Then this operation has suffered a serious intelligence failure, young colleague, and nobody packed snacks for the field team or juice boxes either.” Molly steered Bill toward the refreshment table, where the presents sat beneath a banner proclaiming, Please Take One Cup, Not Everything, in bold red letters. “Perhaps we should reconsider. Stealing from children feels worse when the children are organized, observant, and apparently armed with dinosaurs and excellent questions about everything.” Bill studied the gift mountain, then the boy, then a toddler guarding a glittery scooter with the focused vigilance of a tiny customs officer there. “Fine. We’ll return every present, help with cleanup, and leave before anyone asks why our disguises have price tags from the discount bin still attached.” Molly removed her duck cardigan and offered it to the shivering magician, whose rabbit had returned wearing a bow tie and tremendous self-confidence for once. “This cardigan is lovely, with pockets, but please don’t check them; one contains emergency confetti and the other contains dignity, mostly, on alternate Tuesdays only.” The children applauded as Bill stacked chairs, Molly swept crumbs, and the birthday girl appointed both thieves honorary helpers, pending a thorough background check tomorrow. “Next time, we should rob a bake sale instead. The cookies have no guardians, and our new references say we’re excellent with cleanup and toddlers.”
created_at
"2026-09-23T08:25:24.02Z"
9/23/2026, 8:25:24 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
6194
moderation_latency
(null)
generation_time
58549
tokens_prompt
75
tokens_completion
1387
native_tokens_prompt
66
native_tokens_completion
6502
native_tokens_completion_images
0
native_tokens_reasoning
5891
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
0.0032576
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790151924-msL3iY4JbLgNjvo8de9y"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790151924-x6hmuFaNUmRl1Y7QNc2s"
upstream_id
"resp_06f7fd194b7d0b23016ab38cf419fc87d1b991540eb4a67c01"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_06f7fd194b7d0b23016ab38cf419fc87d1b991540eb4a67c01"
is_byok
false
latency
2838
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0032576
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
100.00%
Matches word count
n/a
neededClean
false
words
500
99.95%
Dialogue to Total Word Ratio
Ratio: 49.90%, Deviation: 0.10%
neededClean
false
wordsTotal
501
wordsDialogue
250
99.9751%