NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36NM9XE194B6KYQJXFA8AJ7
Run Details
Model:
GPT-6 Luna (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
19.4s
Cost:
$0.000625
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
27.17%
Matches word count
(details)
61.07%
Dialogue to Total Word Ratio
(details)
44.1168%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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 crouched behind the garden gnome, which had the smug expression of a statue that knew where the spare key was. Across the lawn, a banner announced ELLIOT’S SEVENTH BIRTHDAY in letters large enough to be read from space. Beneath it, a mountain of wrapped presents guarded the patio door. Bill adjusted his bow tie. “We enter as entertainers, locate the gifts, and leave before anyone asks us to do a magic trick.” “You’re wearing a bow tie with tiny clowns on it,” Molly whispered. “It’s called camouflage.” “It’s called evidence.” From the garden came a shriek of delighted children. A party game had begun, involving a sack, several potatoes, and what sounded like a dispute before an international court. Molly unfolded their plan. It was written on the back of a takeaway menu and featured a map, three arrows, and a drawing of a horse wearing sunglasses. “Why is there a horse?” Bill asked. “I needed something to represent the pony rides.” “They have pony rides?” “They might.” Bill peered through the hedge. “There is a pony.” “Then the plan is already ahead of us.” The pony stood beside the presents, wearing a pink party hat and the air of a retired detective. Bill had once been thrown by a Shetland pony and still referred to the incident as “the fall of the Roman Empire.” Molly tapped the map. “We slip in during the cake ceremony. You distract the adults with your juggling.” “I don’t juggle.” “You said you could.” “I said I could learn.” “How many balls do you have?” “One.” “Then you’ll be juggling one ball in a room full of people. Try not to let them notice.” Bill looked wounded. “My talents are often misunderstood.” The back door opened. A woman carrying a tray of cupcakes hurried past them, leaving it ajar. Molly and Bill exchanged the bright, reckless look of people whose plans had just been improved by circumstance. Inside, the presents glittered beneath a paper sun. Molly reached for the biggest box, then stopped. A card on top read: TO ELLIOT, FROM GRANDMA. LOVE YOU LOTS. Bill leaned close. “That’s inconveniently sweet.” “Don’t get sentimental. We’re professionals.” “Professionals can recognize a strong card.” A small boy appeared, wearing a cardboard crown and clutching a toy dinosaur. “Are you the clowns?” Bill glanced at Molly. “We’re the—” “Clowns,” Molly said, swiftly. “He’s the quiet one.” The boy pointed to the presents. “You can help me open them!” For one breath, the thieves considered the mountain of loot. Then Molly picked up the nearest box and handed it to him. Bill began juggling his single ball, which bounced off his shoe, struck the pony’s hat, and landed neatly in a punch bowl. The children applauded as if he’d performed surgery. Molly grinned. “New plan. We steal the cake.” Bill examined the enormous frosted cake. “At last, a job within my skill set.” They left with two slices, one paper crown, and the pony’s solemn approval. The presents stayed where they belonged, though Bill later insisted the cake had been worth more on the black market.
created_at
"2026-09-23T08:20:32.822Z"
9/23/2026, 8:20:32 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
9187
moderation_latency
(null)
generation_time
19286
tokens_prompt
75
tokens_completion
954
native_tokens_prompt
66
native_tokens_completion
1236
native_tokens_completion_images
0
native_tokens_reasoning
549
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
0.0006246
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790151632-SRzVLFXPb3UgWNy5J0Y7"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790151632-5dqfEbGPgdVIAkYHZ4LD"
upstream_id
"resp_0d086933126a65fd016ab38bd0eab087d1b081895edb868930"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_0d086933126a65fd016ab38bd0eab087d1b081895edb868930"
is_byok
false
latency
451
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0006246
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
27.17%
Matches word count
n/a
neededClean
false
words
519
61.07%
Dialogue to Total Word Ratio
Ratio: 33.14%, Deviation: 3.14%
neededClean
false
wordsTotal
519
wordsDialogue
172
44.1168%