NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX40G0HA8J91M07Q9EVGC2KT
Run Details
Model:
GPT-5.6 Luna (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
7.2s
Cost:
$0.006318
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
7.43%
Dialogue to Total Word Ratio
(details)
3.7165%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% 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, peering through a gap in the leaves at the brightly decorated house. Balloons bobbed above the fence, a cardboard dragon guarded the front door, and somewhere inside, children shrieked with the terrifying enthusiasm of people who had consumed too much cake. Bill adjusted his black woolly hat, which had a pom-pom on top. “Your disguise is undermined by the pom-pom,” Molly whispered. “It’s winter-themed,” Bill said. “It’s July.” “Exactly. No one suspects seasonal confusion.” Molly unfolded a hand-drawn map across the hedge. It was covered in arrows, circles, and one large question mark pointing at the refreshments table. “The presents are stacked beside the living room window,” she said. “We slip through the side gate, cross the patio, enter through the unlocked laundry door, and fill the van.” “Which van?” “The van.” “The van with the words ‘Molly and Bill’s Totally Normal Florist Service’ painted on the side?” “It’s a very convincing cover.” “It has a giant cartoon cabbage wearing sunglasses.” “Children love cabbages.” “Children hate cabbages. Adults hate cabbages. Cabbages probably hate cabbages.” A shriek rose from the house, followed by enthusiastic clapping. Molly glanced at her watch. “The magician is performing. That gives us eight minutes.” “Eight minutes to steal every present?” “Seven minutes now.” Bill looked through the hedge. “There are at least thirty children.” “They’re distracted.” “By what?” “Magic.” “Magic is unreliable. I once saw a magician pull a rabbit from a hat, and the rabbit looked furious.” Molly tucked the map into her coat. “You’re overthinking this.” “I’m underthinking the escape route. There’s a bouncy castle blocking the driveway.” “We go over it.” “With the presents?” “Use your imagination.” Bill sighed and lifted the hedge branch. “Fine. Operation Birthday Bandit begins.” They slipped through the gate, stepped across the patio, and reached the laundry door. Molly tried the handle. It opened immediately. Bill blinked. “That was alarmingly easy.” “Criminals appreciate hospitality.” Inside, they crept down a hallway decorated with paper stars. A small boy in a pirate costume emerged from the kitchen carrying a plate of sausages. He stared at them. Molly froze. Bill raised one finger to his lips. “We’re the present inspectors.” The boy narrowed his eyes. “You don’t look like inspectors.” “That’s because we’re undercover,” Molly said. “Are you pirates?” “Technically,” Bill answered. The boy pointed toward the living room. “The presents are there.” “Thank you,” Molly said. “And the grown-ups are playing musical chairs.” “Excellent.” “But you have to say the password.” Molly leaned closer. “What’s the password?” “Banana.” “Banana,” she said. The boy shook his head. “Wrong.” Bill whispered, “You just asked him.” The boy grinned. “That wasn’t the password.” A door opened behind them. Someone shouted, “Who wants more jelly?” Molly grabbed Bill’s sleeve. “Plan B.” “What’s Plan B?” “Become entertainment.” They burst into the living room, where every child turned toward them. Molly seized a paper crown, placed it on Bill’s head, and announced, “Behold! The birthday thieves!” The children cheered. Bill stared at her. “That is not better!” Within seconds, the children surrounded them, demanding tricks, juggling, and a demonstration of Bill’s famous disappearing act. Molly shoved him toward the cake while she carried the presents outside, one at a time, until a little girl caught her. “That’s my teddy bear,” the girl said. Molly looked at the enormous pile in her arms. “It’s a very advanced inspection.” The girl hugged the bear. “You can inspect mine after cake.” Molly glanced at the children, the glittering balloons, and Bill attempting to juggle three cupcakes. She sighed, returned the gifts, and joined the party. By sunset, they had stolen exactly nothing except two slices of cake and a paper napkin. Bill licked frosting from his thumb. “Best heist ever?” Molly nodded. “We got away with dessert.”
created_at
"2026-07-09T17:58:32.242Z"
7/9/2026, 5:58:32 PM
model
"openai/gpt-5.6-luna-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1738
moderation_latency
(null)
generation_time
6566
tokens_prompt
77
tokens_completion
1336
native_tokens_prompt
66
native_tokens_completion
1042
native_tokens_completion_images
0
native_tokens_reasoning
140
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
0.006318
router
(null)
provider_responses
0
endpoint_id
"eebea444-34a5-4642-a766-cb319471d33a"
id
"resp_0fc2cc306a4ae718016a4fe148503c8193bcc46d46b3d1f161"
is_byok
false
latency
267
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-1783619912-6ImWHwcB5kOC9Xyf5ftQ"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783619912-dmfKo17uOX9DPALxYLkS"
upstream_id
"resp_0fc2cc306a4ae718016a4fe148503c8193bcc46d46b3d1f161"
total_cost
0.006318
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
633
7.43%
Dialogue to Total Word Ratio
Ratio: 42.79%, Deviation: 7.21%
neededClean
false
wordsTotal
638
wordsDialogue
273
3.7165%