NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36NTG1S6F5E1CCQ7JMZ6TJD
Run Details
Model:
GPT-6 Luna (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
15.1s
Cost:
$0.000569
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
99.98%
Matches word count
(details)
0.65%
Dialogue to Total Word Ratio
(details)
50.3157%
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
“Tell me again why the presents are in the room marked ‘Presents,’” Molly whispered. “Because the host has a gift for labeling,” Bill whispered back. “A gift we intend to redistribute.” “To whom?” “To ourselves.” “That’s not redistribution. That’s re-gifting with extra steps.” They crouched behind a cardboard castle in the community hall, where a banner proclaimed HAPPY BIRTHDAY, NED! in letters that drooped like they’d heard the plan. Beyond it, children shrieked around a cake shaped like a dinosaur. Bill wore a paper crown. Molly wore two, having misunderstood the costume instructions. “Right,” Bill said. “We wait until the children are distracted, then take every present.” “Every present?” “Every single one.” “Bill, there are at least thirty.” “Thirty-two. I counted.” “You counted children’s presents?” “I counted bows. It’s called reconnaissance.” “It’s called sitting in a shrubbery with binoculars.” “I was committed.” Molly peered past the castle. A small boy was trying to feed a balloon to the dinosaur cake. “This is a terrible idea.” “You said that about the museum.” “And the museum had alarms.” “This has a clown.” “Exactly. Clowns are unpredictable.” A balloon popped. Bill ducked so sharply his crown slid over one eye. “Steady,” Molly said. “We’re professionals.” “Professionals don’t wear two crowns.” “I’m the queen. You’re the jester.” “I thought I was the getaway driver.” “You’re also the jester.” A woman with a clipboard approached. “You two are the entertainers, right?” Molly glanced at Bill. “We’re—” “Absolutely,” Bill said. “The entertainment.” The woman handed them a sack of party favors. “Great. Hand these out, then lead the children in a treasure hunt.” She hurried away. Bill examined the sack. “This is either an opportunity or a trap.” “It’s party favors.” “Exactly what they’d use as bait.” Molly sighed. “We hand these out, lead the hunt, and then leave.” “What about the presents?” “We leave them.” “Even the large one?” “Especially the large one. It’s labeled ‘From Grandma.’” Bill looked stricken. “Grandma’s involved?” “She gave it. That’s usually how presents work.” He watched the children gather for the treasure hunt. “If we guide them away from the presents, we could—” “No.” “—make sure they find every clue.” “No.” “—and perhaps discover a clue pointing toward the cake.” “Bill.” “Fine. I’ll be noble.” They passed out party favors. Bill gave one child three and another a lecture about fair distribution. Molly led the hunt, which ended beneath the refreshment table, where the children found a paper crown and declared her queen. “Your Majesty,” Bill said, bowing. “Shall we seize the presents?” “No,” Molly said. “We shall seize the leftover cake.” “That’s not a heist.” “It is if you take the last slice.” Bill considered this. “And the presents?” “Let the birthday boy open them.” Ned tore the paper from a box and gasped at a red bicycle. Bill watched, blinking. “That seems… nice.” “Careful,” Molly said. “You’re developing a conscience.” “Can we steal a party favor?” “Bill, you’ve already got three.” He looked into his sack. “Then we’re rich.”
created_at
"2026-09-23T08:23:55.711Z"
9/23/2026, 8:23:55 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
6094
moderation_latency
(null)
generation_time
15089
tokens_prompt
75
tokens_completion
1064
native_tokens_prompt
66
native_tokens_completion
1125
native_tokens_completion_images
0
native_tokens_reasoning
393
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
0.0005691
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790151835-d1k5tXA28g8ywMNldomi"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790151835-vagVle8JzsuMFsMiqJCa"
upstream_id
"resp_0c1df930f62a8733016ab38c9bcf1487d184382f00c926e9d0"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_0c1df930f62a8733016ab38c9bcf1487d184382f00c926e9d0"
is_byok
false
latency
645
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0005691
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
502
0.65%
Dialogue to Total Word Ratio
Ratio: 59.96%, Deviation: 10.04%
neededClean
false
wordsTotal
502
wordsDialogue
301
50.3157%