NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ0MZN16KJ9Q0BPF72H5BKFA
Run Details
Model:
GPT-5.5
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
33.3s
Cost:
$0.033390
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
65.41%
Dialogue to Total Word Ratio
(details)
32.7026%
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 the birthday invitation flat on the diner table and pinned its corners with four packets of ketchup. “Seven years old,” Bill said, squinting. “That’s a lot of candles for a mark.” “It’s a lot of presents for us,” Molly said. “Look at the address. Big house, bigger lawn, inflatable castle. People who rent a castle for a child don’t wrap socks.” Bill stirred his coffee with a french fry. “I have moral reservations.” “You once stole a priest’s bicycle.” “It had a bell shaped like a duck. That was destiny. This is different. Children cry loudly.” “Then we won’t make them cry. We’ll take the presents before anyone opens them.” “That is exactly what makes them cry.” Molly sighed. “Fine. We don’t take all the presents.” “You said all the presents.” “I was being dramatic. We take the expensive-looking ones and leave educational toys.” Bill nodded. “So we’re the good guys.” “We’re the literacy bandits.” A waitress passed, glanced at the invitation, and said, “Cute. Your kid’s party?” “Our niece,” Molly said. “Our accountant,” Bill said at the same time. The waitress paused. Molly smiled. “She’s very advanced.” When the waitress left, Bill leaned in. “We need disguises.” “Already handled.” Molly pulled a flyer from her bag. “They hired Princess Petunia and Captain Sprinkle.” Bill read it upside down. “Captain Sprinkle sounds like a weather event at a bakery.” “You’ll be him.” “I don’t do tights.” “You do if they come with a cape.” “I do capes,” Bill conceded. “Capes have authority.” “I’ll be Princess Petunia. We arrive at two, entertain the children, locate the gift table, and at two-thirty we perform the Great Vanishing Act.” “Where the presents vanish?” “No, Bill, where we vanish. With selected presents.” “What if the real Princess Petunia arrives?” “She won’t. I called and canceled.” “You canceled a princess?” “I said the birthday boy had become allergic to royalty.” Bill’s face brightened. “That’s good. Specific. Medical.” “Thank you.” “What about Captain Sprinkle?” “I said he’d been detained by drizzle.” Bill tapped the table. “You’re wasted in crime. You should run a school.” “That is crime with paperwork.” A toddler at the next booth dropped a spoon and screamed as though betrayed by gravity. Bill flinched. “See?” he whispered. “They have alarms built in.” “We’ll distract them with balloon animals.” “Can you make balloon animals?” “I can make balloon questions.” “What’s a balloon question?” “A long squeaky thing that makes people wonder if it’s a dog.” Bill considered this. “Children are forgiving if there’s cake.” “Exactly.” “And parents are distracted if there’s chaos.” “Exactly.” “And we are experts in chaos.” “Bill, you once got trapped in a revolving door for nineteen minutes.” “I was studying rotation.” Molly slid a crayon map across the table. “Entry through the side gate. Gift table here, beside the patio. Kitchen here. Escape route through the hedge.” Bill studied the map. “Why is there a dragon?” “That’s the barbecue.” “And these tiny knives?” “Candles.” “Your cartography is hostile.” “Your face is hostile, but I still work with it.” Bill pointed. “What’s our carrying capacity?” “One rolling magician’s trunk.” “Do we have one?” “We will.” “Are we stealing that too?” “Borrowing from a magician who owes me eighty dollars.” “How does a magician owe you money?” “He made my patience disappear.” Bill smiled despite himself. “What if the children want autographs?” “Sign ‘Princess Petunia.’” “I’m Captain Sprinkle.” “Then sign that.” “I’m not comfortable forging meteorological authority.” Molly folded the invitation. “Focus. We get in, we charm, we lift, we leave.” “With only the fancy presents.” “Yes.” “No teddy bears.” “No teddy bears.” “No handmade cards.” “No handmade cards.” “No boxes that rattle sadly.” “Agreed.” Bill sat back. “I want a code word.” “Fine. ‘Cupcake.’” “That’s too common at a party.” “‘Bankruptcy.’” “That’s too uncommon at a child’s party.” “Not if the parents see the clown invoice.” Bill laughed, then caught himself. “Molly, suppose the birthday boy is nice?” “Then we leave him something excellent.” “Like what?” She reached into her bag and pulled out the duck-shaped bicycle bell. It quacked when she squeezed it. Bill gasped. “You kept it?” “For emergencies.” “That bell changed me.” “It can change him too.” Bill grew solemn. “All right. We rob the party humanely.” Molly raised her coffee cup. “To humane robbery.” Bill clinked his mug against hers. “To literacy, capes, and selective villainy.” At the counter, the waitress shouted, “Anyone here order the birthday pancake?” Molly and Bill looked at each other. Bill whispered, “Reconnaissance?” Molly grabbed the invitation. “Research.” They slid from the booth with the dignity of professionals and the speed of people chasing whipped cream.
created_at
"2026-04-24T21:05:27.597Z"
4/24/2026, 9:05:27 PM
model
"openai/gpt-5.5-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1013
moderation_latency
(null)
generation_time
33180
tokens_prompt
77
tokens_completion
1562
native_tokens_prompt
66
native_tokens_completion
1102
native_tokens_completion_images
0
native_tokens_reasoning
0
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"
usage
0.03339
router
(null)
provider_responses
0
endpoint_id
"58e5b336-423e-430b-a2ab-8bc353f0c51b"
id
"resp_0f1dcedc2d6a42520169ebdb17ab788194903d632261356a31"
is_byok
false
latency
279
model_permaslug
"openai/gpt-5.5-20260423"
provider_name
"OpenAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777064727-KHRTS5L8REocnDgs3ZKt"
session_id
(null)
api_type
"completions"
id
"gen-1777064727-0cbvJRIY7FGR0P44BkWB"
upstream_id
"resp_0f1dcedc2d6a42520169ebdb17ab788194903d632261356a31"
total_cost
0.03339
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
776
65.41%
Dialogue to Total Word Ratio
Ratio: 72.91%, Deviation: 2.91%
neededClean
false
wordsTotal
779
wordsDialogue
568
32.7026%