NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36R183D1P47R8ACW5A1CNF9
Run Details
Model:
GPT-6 Luna (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
19.2s
Cost:
$0.000708
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.45%
Dialogue to Total Word Ratio
(details)
0.2260%
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
Rain ticked against the bakery awning as Molly unfolded a map of the community hall across a crate of stale cinnamon rolls. Bill leaned over it, wearing a black turtleneck and the expression of a man who had mistaken confidence for competence. “Right,” Molly said. “The birthday party starts at two. We enter, locate the presents, and leave with every last one.” “Define ‘every last one.’” “All of them, Bill.” “What about the enormous one?” “Especially the enormous one.” He studied the map. “That’s a cake.” “It’s drawn like a present.” “It has candles.” “Children are unpredictable. Perhaps they gift each other cake.” Bill nodded gravely. “A dangerous custom.” Molly tapped the hall. “We’ll blend in.” “In what?” “Party clothes.” “I own a tie.” “You own a tie shaped like a fish.” “It’s festive.” “It has a hook through its forehead.” “Also festive.” Molly sighed and rolled up the map. “We’ll go as entertainers. I’ll be a magician; you’ll be my assistant.” “Why am I the assistant?” “Because you can’t do magic.” “I once made a sandwich disappear.” “You ate it.” “Exactly. No witnesses.” They crossed the parking lot toward the hall. Through the windows, children darted beneath paper streamers while a woman arranged gift bags on a table. Bill peered in. “Look at that mountain of presents. We could retire.” “We could buy a house.” “With a room just for presents?” “No.” “Then I’m not retiring.” Molly adjusted the purple top hat she’d borrowed from a costume shop. “Remember, we’re charming. We’re harmless. We’re here to entertain.” “I can do charming.” “Last week you told a cashier her smile looked expensive.” “It did. Very high-quality smile.” At the entrance, a small boy opened the door and stared up at them. “Are you the magician?” Molly smiled. “We certainly are.” “Can you make my dad disappear?” Bill crouched beside him. “Can you make him disappear first? Then I’ll study the technique.” “Bill.” “What? It’s a professional question.” The boy grabbed Molly’s hand and tugged them inside. “Come on! Everyone’s waiting!” Molly whispered, “We go in, we assess the presents, we improvise.” “Could we improvise a quick exit?” “Not before the cake.” A woman in a paper crown hurried over. “Wonderful, you’re here! The children are ready. Could you help me bring the presents to the gift table?” Bill glanced at Molly. “This is either ideal or a trap.” “It’s a children’s party. The trap is probably glitter.” They carried the presents across the room, stacking them neatly beneath a banner that read HAPPY BIRTHDAY, LEO. Every child watched, delighted. “Careful with the blue one,” Leo announced. “That’s for me.” “Of course,” Molly said. “And the red one.” “Also yours?” “No, that’s for my sister. She’s at home.” Bill froze. “Your sister isn’t here?” “She’s sick.” Molly looked at the pile, then at Bill. “We’re not stealing these.” “I know,” he murmured. “I’m deciding whether to steal the cake instead.” A child handed him a paper crown. Bill put it on. “You look like the king of bad ideas,” Molly said. “Then kneel, assistant.” The children cheered as Molly pulled a rubber chicken from her hat. Bill bowed, and the crown slid over his eyes. “Behold,” Molly announced, “the disappearing thief!” “Where?” shouted Leo. Bill shuffled blindly toward the cake. “I’m right here, and I’m beginning to see why professionals bring maps.”
created_at
"2026-09-23T09:02:34.106Z"
9/23/2026, 9:02:34 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
4612
moderation_latency
(null)
generation_time
19184
tokens_prompt
75
tokens_completion
1377
native_tokens_prompt
66
native_tokens_completion
1402
native_tokens_completion_images
0
native_tokens_reasoning
605
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
0.0007076
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790154154-qq8o4CvhM37nNGTvgang"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790154154-xTOYERptEMaOW6vZWiSK"
upstream_id
"resp_02ad967be2b8ae56016ab395aa36f887d1abc7eadb9bb096af"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_02ad967be2b8ae56016ab395aa36f887d1abc7eadb9bb096af"
is_byok
false
latency
964
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0007076
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
561
0.45%
Dialogue to Total Word Ratio
Ratio: 59.61%, Deviation: 10.39%
neededClean
false
wordsTotal
562
wordsDialogue
335
0.2260%