NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36R5HRF3NP3W1M6JX07HH9G
Run Details
Model:
GPT-6 Luna (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
18.7s
Cost:
$0.000735
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
95.45%
Dialogue to Total Word Ratio
(details)
47.7235%
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
Molly spread a napkin across the café table. On it she had drawn a lopsided cake, three balloons, and a question mark wearing a hat. “Is that the party hall?” Bill asked. “It’s the party hall.” “Why is it shaped like a boot?” “That’s the cake.” Bill leaned closer. “Then what’s the boot?” “The party hall.” He nodded. “Good. We’ll need a better map.” “We don’t need a map. We need a plan.” “Those are usually the same thing, except one has more arrows.” Molly tapped the question mark. “The presents will be here.” “How do you know?” “Because children put presents beside cakes.” “Children put jam in their hair. I wouldn’t base a criminal enterprise on their customs.” “Fine. We enter the party, locate the gifts, and leave with every last one.” “Every last one?” Bill frowned. “That sounds greedy.” “We’re thieves, Bill.” “I thought we were specialists in temporary borrowing.” “You once borrowed a bicycle and mailed it back in pieces.” “It was a folding bicycle.” Molly folded the napkin in half. “We need disguises.” “I can be a magician.” “You can’t do magic.” “I can make a coin disappear.” “Where does it go?” “Into my sleeve.” “Then you can make a coin appear in your sleeve. That’s not the same.” “I’ll need a longer sleeve.” Molly sighed. “Perhaps we go as entertainers.” “I can juggle.” “You dropped three oranges into a fountain.” “They were slippery.” “They were oranges.” “I can juggle balloons.” “Everyone can juggle balloons. They’re balloons.” “Exactly. I’ll blend in.” Molly stared at the napkin. “Then, once the children are distracted—” “With balloons?” “With the cake.” “Are we stealing the cake too?” “No.” “Because that’s the sort of detail a plan should settle early.” “We take the presents and go.” “What if there’s a gift for us?” “There won’t be.” “What if there’s a card addressed to ‘Two Handsome Gentlemen’?” “Then you may keep the card.” Bill sat back, offended. “I’m not doing this for cards.” A little girl at the next table was carefully wrapping a cardboard box in newspaper. She tied it with string, then wrote something on a tag in large, uneven letters. Bill glanced over. “What’s she making?” “Probably a present.” “For the party?” Molly looked at the drawing on their napkin: a cake, balloons, a question mark in a hat. For a moment, the plan seemed less like a plan and more like two grown adults trying to outwit a room full of children. “What would you do with all those presents?” she asked. “Sell them.” “To whom?” Bill paused. “People who like presents.” “People who like presents usually prefer choosing their own.” “Then we’d… redistribute them.” “By keeping the money?” “Redistribution is complicated.” Molly stood. “We’re not stealing from children.” Bill rose too. “I was just about to say the same thing.” “You were not.” “I was approaching it from the other direction.” They left a few coins on the table. Bill picked up the napkin. “Don’t,” Molly said. “I’m saving the map.” “It’s a cake.” “Could be useful. We’re going to the party, aren’t we?” “To return the napkin?” “To bring a gift.” Molly eyed him. “You have a gift?” Bill produced a coin from his sleeve. She looked at it. “Fine,” he said. “We’ll buy something.”
created_at
"2026-09-23T09:04:55.075Z"
9/23/2026, 9:04:55 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
4388
moderation_latency
(null)
generation_time
18592
tokens_prompt
75
tokens_completion
1369
native_tokens_prompt
66
native_tokens_completion
1457
native_tokens_completion_images
0
native_tokens_reasoning
688
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
0.0007351
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790154295-faVarE8eLKvISpGOFy8i"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790154295-Ezjvbe1BilzHPAgYuxZz"
upstream_id
"resp_054a602d33470657016ab39637374887d1a69561a88032958e"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_054a602d33470657016ab39637374887d1a69561a88032958e"
is_byok
false
latency
861
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0007351
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
549
95.45%
Dialogue to Total Word Ratio
Ratio: 69.03%, Deviation: 0.97%
neededClean
false
wordsTotal
549
wordsDialogue
379
47.7235%