NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36R7DJX34R146CB4E6JQ6MP
Run Details
Model:
GPT-6 Luna (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
16.8s
Cost:
$0.000647
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
99.98%
Dialogue to Total Word Ratio
(details)
49.9923%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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
The map of the neighborhood was spread across the diner table, pinned beneath a milkshake, a ketchup bottle, and Bill’s left elbow. “According to this,” said Molly, “the birthday party is at the house with the enormous yellow balloon.” “Could be any house,” said Bill. “It says ‘HAPPY BIRTHDAY, LEO’ on the balloon.” “Then it’s probably the one with Leo.” Molly tapped the map. “We arrive, blend in, locate the presents, and leave with the entire pile.” “Do we have a pile-shaped vehicle?” “We have a car.” “Our car has one working door and a boot full of novelty mustaches.” “Then we’ll wear the mustaches. Nobody suspects a mustache.” “Nobody suspects a mustache because nobody wears one indoors.” “Exactly. They’ll be too confused to stop us.” Bill studied the map upside down. “What’s the code name for this operation?” “Operation Birthday Bandit.” “Too obvious. How about Operation Auntie June?” “Why Auntie June?” “Because every party has an Auntie June. She knows where the plates are, and people trust her.” Molly considered him. “Do you know how to be Auntie June?” “I can say, ‘Where’s the nice sharp knife?’” “That’s not a personality. That’s a warning.” A waitress set down two coffees. Bill thanked her, then leaned closer to the map. “What if the presents are already guarded?” “By whom?” “Children.” “Children are small. We’re professionals.” “Children are also loud, unpredictable, and sticky.” Molly glanced at the smudge of jam on his sleeve. “So are you.” “I’m sticky in a controlled way.” She pulled a notepad from her coat. “We need a distraction.” “I can juggle.” “You dropped three oranges and a shoe last time.” “The shoe was part of the act.” “It belonged to the grocer.” Bill frowned at the map. “What if we become entertainers? Nobody questions entertainers.” “Nobody questions entertainers because they’re waiting for them to stop.” “I could do magic.” “Can you make a present disappear?” “I can make a birthday cake disappear.” “That’s not magic. That’s why we’re banned from the bakery.” Bill brightened. “We could take the cake instead. Less wrapping.” “We are stealing presents, Bill. That’s the job.” “Right. Presents. Do we know what’s in them?” “No.” “What if they’re socks?” “Then we’ll be criminals with warm feet.” A gust of wind rattled the diner window. On the sidewalk outside, a little boy marched past carrying a paper crown. He was followed by a woman carrying a gift bag nearly as large as he was. Bill watched them go. “That bag could hold a pony.” “Or a toy pony.” “Either way, awkward getaway.” Molly folded the map. “We enter as guests, act natural, and don’t touch anything until we’ve assessed the situation.” “Natural,” Bill repeated, attempting a smile so strained it looked like a dental emergency. “And if a child asks who we are?” “Uncle Bill and Auntie Molly.” “We’re not related.” “Neither are most uncles at parties.” Molly stood. “Fine. Operation Auntie June begins at noon.” Bill put on one of the novelty mustaches. It covered his mouth. “Can you still talk?” “Mm-hm.” “Good. Practice saying ‘happy birthday’ without frightening anyone.” He tried. The mustache slid into his coffee. Molly sighed. “We’re going to need a better disguise.” “And a towel,” said Bill.
created_at
"2026-09-23T09:05:56.332Z"
9/23/2026, 9:05:56 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
7945
moderation_latency
(null)
generation_time
16758
tokens_prompt
75
tokens_completion
1094
native_tokens_prompt
66
native_tokens_completion
1280
native_tokens_completion_images
0
native_tokens_reasoning
516
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
0.0006466
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790154356-9Y4SQYsrfnV4fGeOIZ6u"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790154356-zGV5ZiyG1kgbUvFTrmAA"
upstream_id
"resp_00c1d520be37ee3b016ab396747b3887d1965801d885b9018e"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_00c1d520be37ee3b016ab396747b3887d1965801d885b9018e"
is_byok
false
latency
530
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0006466
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
0.00%
Matches word count
n/a
neededClean
false
words
540
99.98%
Dialogue to Total Word Ratio
Ratio: 70.06%, Deviation: 0.06%
neededClean
false
wordsTotal
541
wordsDialogue
379
49.9923%