NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX40XH8RPYTA4ECGT1J851HC
Run Details
Model:
GPT-5.6 Luna (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
6.5s
Cost:
$0.005844
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
85.11%
Dialogue to Total Word Ratio
(details)
42.5548%
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 hedge, wearing a black turtleneck, a black balaclava, and the expression of someone who had just discovered a suspiciously large amount of glitter in her soup. Beside her, Bill peered through binoculars at the birthday party next door. The garden was decorated with balloons, paper crowns, and a banner reading HAPPY BIRTHDAY, HAROLD! One corner contained a mountain of brightly wrapped presents. Another contained twelve children, three exhausted adults, and a clown who appeared to be reconsidering every decision that had led him there. “Target identified,” Bill whispered. “Present pile, north side of the patio.” “Any security?” “Two parents, one grandmother, a terrier, and a magician.” “Which is the greatest threat?” “The magician. He’s already made a rabbit disappear.” Molly lowered her binoculars. “A rabbit?” “Possibly a handkerchief. I couldn’t tell. There was screaming.” “Excellent. We go in during the cake.” “Why the cake?” “Because everyone looks at the cake.” “Everyone looks at the cake when it’s being carried in. During the eating, everyone looks at the children.” Molly considered this. “Then we go in during the singing.” “Everyone looks at the birthday child.” “Bill, you’re making this unnecessarily difficult.” He unfolded their blueprint, which was actually a takeaway menu with arrows drawn across the section marked FAMILY COMBOS. “We enter through the side gate, distract the terrier with sausages, cross the lawn beneath the balloon arch, and remove the presents using the laundry trolley.” Molly glanced at the trolley. It had one wheel missing and a Christmas wreath hanging from its handle. “It’s festive,” Bill said. “It’s June.” “Festive is an attitude.” A shriek rose from the garden. The magician had produced a dove from somewhere, and the children were chasing it toward the hedge. Molly and Bill flattened themselves into the shrubbery as the bird fluttered overhead. “Now?” Bill asked. “Now.” They slipped through the side gate with the careful silence of professionals, though Bill’s trousers snagged on the latch and made a noise like a startled goose. The terrier charged. Molly tossed a sausage. The dog caught it, swallowed it, and immediately returned for a written explanation. They reached the patio just as the adults began singing “Happy Birthday.” Every child turned toward Harold, who sat on a throne made from stacked cushions. The magician stood behind him, holding a cake that was visibly smoking. Molly seized the nearest present. Bill grabbed two more. “Careful,” she hissed. “No crushing.” “I’m treating them like eggs.” “You once dropped an egg in a getaway car.” “It was a hostile egg.” The balloon arch collapsed with a soft, festive sigh. Everyone turned. Molly and Bill froze beside the present pile, each holding a brightly wrapped box. Harold stared at them from his cushion throne. “Are you the surprise entertainment?” he asked. Bill slowly lifted the box. “Yes.” “What do you do?” Molly smiled. “We make presents disappear.” Harold clapped. The children cheered. Five minutes later, Molly and Bill were trapped inside a ring of delighted children, performing the worst magic act in recorded history. Every present vanished into their laundry trolley, only to be returned dramatically to the birthday table. By the end, Harold had received everything, including a new bicycle, three action figures, and the terrier, which had climbed into the trolley and refused to leave. Molly and Bill escaped over the hedge empty-handed. “Successful?” Bill asked. Molly brushed glitter from her sleeve. “We stole the party.” “Worth anything?” “Not financially.” Behind them, the children applauded. The magician took a bow. The terrier barked. Bill sighed. “Next time, we rob a bank.” “Too much security,” Molly said. “And probably fewer sausages.”
created_at
"2026-07-09T18:05:55.371Z"
7/9/2026, 6:05:55 PM
model
"openai/gpt-5.6-luna-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1683
moderation_latency
(null)
generation_time
6195
tokens_prompt
77
tokens_completion
1233
native_tokens_prompt
66
native_tokens_completion
963
native_tokens_completion_images
0
native_tokens_reasoning
139
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
0.005844
router
(null)
provider_responses
0
endpoint_id
"eebea444-34a5-4642-a766-cb319471d33a"
id
"resp_073d34a88c5a204c016a4fe303727c81959d9aea7aaf2d198c"
is_byok
false
latency
246
model_permaslug
"openai/gpt-5.6-luna-20260709"
provider_name
"OpenAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1783620355-6FCnrC8upb0zsBxaI3l6"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783620355-ot4iMca7yByWaZkDYiYD"
upstream_id
"resp_073d34a88c5a204c016a4fe303727c81959d9aea7aaf2d198c"
total_cost
0.005844
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
606
85.11%
Dialogue to Total Word Ratio
Ratio: 31.80%, Deviation: 1.80%
neededClean
false
wordsTotal
607
wordsDialogue
193
42.5548%