NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ0PK03BW9YEQP8JXH5HKJX3
Run Details
Model:
GPT-5.5 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
25.1s
Cost:
$0.030270
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
96.27%
Dialogue to Total Word Ratio
(details)
48.1327%
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 unfolded the crayon map across the bonnet of Bill’s van. It showed a bouncy castle, a cake table, and something labelled PRESENT MOUNTAIN in green glitter. Bill squinted. “Why is the dog wearing a crown?” “That’s not the dog. That’s the birthday boy.” “Small lad.” “He’s six.” “Ah. I was six once.” “Hard to imagine.” “I had dreams, Molly.” “You had a criminal record in pencil.” Bill tapped the glittery mountain. “So the presents are here, under the suspiciously cheerful balloon arch.” “Correct.” “And the adults will be distracted by games, cake, and that clown who smells of petrol.” “Correct again.” “And we waltz in, disguised as entertainers.” Molly held up two costumes. One was a unicorn with a sagging horn. The other was a pirate with a moustache already attached. Bill frowned. “I don’t want to be the unicorn.” “You’re the unicorn.” “I have dignity.” “You have a van with ‘Bill’s Discount Ferrets’ painted on the side.” “That was a legitimate business.” “You owned one ferret, and it bit a priest.” “It was a misunderstanding.” Molly shoved the unicorn head at him. “Put it on.” Bill peered into its hollow face. “It looks like it’s seen things.” “It has. Last Tuesday we used it to sneak into a hen do.” “That was not sneaking. They cheered.” “They thought you were a surprise.” “I was.” Molly checked her watch. “Right. We go in at three-fifteen, during musical statues.” Bill nodded gravely. “Children frozen. Adults filming. Perfect.” “No shoving, no frightening anyone, no ruining the cake.” Bill gasped. “What kind of monster do you take me for?” “The kind who ate evidence.” “It was a cupcake wrapper.” “With the cupcake still in it.” “Details.” A shriek of laughter floated from the garden beyond the hedge. Balloons bobbed above it, bright and innocent. Bill shifted his weight. “Molly.” “What?” “Are we certain about this?” She looked up. “About what?” “Stealing presents from children.” “That is the job.” “Yes, but when you say it plainly, it sounds less like a heist and more like being hissed at by nuns.” Molly sighed. “They’re rich children.” “They’re still children.” “They won’t miss one or two presents.” “You said all the presents.” “That was before you developed a conscience in a unicorn head.” Bill hugged the head to his chest. “Maybe I always had one.” “You once stole a wheelbarrow from a scarecrow.” “It looked abandoned.” “It was in a field.” “So was I, emotionally.” Molly pinched the bridge of her nose. “Bill, we planned for three weeks.” “We drew a map in crayon.” “It was themed.” “We learned balloon animals.” “You learned one balloon sausage.” “It brought joy.” A little girl’s voice rang from the garden. “Mummy, the presents are for the hospital after! Remember?” Molly froze. Bill froze harder. Bill whispered, “Hospital?” Molly folded the map very slowly. “Apparently.” “Children’s hospital?” “Sounds like it.” “So present mountain is charity mountain.” “It would appear so.” Bill nodded. “New plan.” Molly raised an eyebrow. “Which is?” “We go in as entertainers.” “And?” “We entertain.” Molly stared at him. “For free?” “For redemption.” “That’s not a currency.” “It is with nuns.” A balloon burst. Bill jumped and dropped the unicorn head. Molly picked it up and shoved it back at him. “Fine. We entertain. But if any child asks for a giraffe, you’re making the sausage and calling it a short giraffe.” Bill pulled on the head. His voice came muffled and heroic. “Molly?” “Yes?” “After this, can we steal something from someone awful?” “Absolutely.” “Like that clown?” “Especially that clown.”
created_at
"2026-04-24T21:33:30.101Z"
4/24/2026, 9:33:30 PM
model
"openai/gpt-5.5-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
5951
moderation_latency
(null)
generation_time
25023
tokens_prompt
77
tokens_completion
1306
native_tokens_prompt
66
native_tokens_completion
998
native_tokens_completion_images
0
native_tokens_reasoning
151
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
0.03027
router
(null)
provider_responses
0
endpoint_id
"58e5b336-423e-430b-a2ab-8bc353f0c51b"
id
"resp_01afc14b3418e2e50169ebe1aa337c81979364677a3c82475e"
is_byok
false
latency
234
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-1777066410-abb1wQpQF3fwCEZw3cth"
session_id
(null)
api_type
"completions"
id
"gen-1777066410-O4odpyd50aO9bfgzoJcf"
upstream_id
"resp_01afc14b3418e2e50169ebe1aa337c81979364677a3c82475e"
total_cost
0.03027
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
595
96.27%
Dialogue to Total Word Ratio
Ratio: 69.13%, Deviation: 0.87%
neededClean
false
wordsTotal
596
wordsDialogue
412
48.1327%