NC Bench
Overview
About
Tests
Models
Model
GPT-4.1
Test
Dialogue tags
Scenario
Write 200 words with 90% dialogue
Run ID
2025-04-14-01JRTSXET9181DEJJ89ERA7B6J
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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
"Bill, do you have the clown nose?" "Of course, Molly. How else are we getting past the balloon arch, brute force?" "I'd actually considered it. But then I remembered you’re allergic to latex." "One time! I sneezed on a magician’s rabbit, and suddenly I’m 'unpredictable.'" "So, you’ve got the nose. Did you bring the decoy gift bag?" "Here. Blue wrapping. Glitter bows. Heavy enough to pass as a box of Legos." "What’s inside?" "Bricks. Literal bricks. Don’t drop it on your toes." "Fine, fine. Now, the plan?" "We mingle. Smile at parents, compliment the cake. You work the snacks table, I scope the present pile." "I heard Little Tommy’s aunt brings gold coins." "Real gold? People still do that?" "I checked her Instagram. #GenerousAuntMarge." "Perfect. Party hats—yes or no?" "Of course yes. Nothing says ‘I belong here’ like paper hats shaped like pirate ships." "If a kid asks who we’re here for?" "We’re friends of…um…Mopsy?" "Mopsy?" "It’s the clown. The hired entertainment. If things go south—juggling distraction." "Remind me why we’re not just robbing banks?" "Fewer laser grids, more cake." "Fair point. Ready, Molly?" "Ready, Bill. Grab the noisemaker. If anyone gets suspicious, you know what to do." "Run and blame the magician?" "Exactly."
id
"gen-1744656317-50sRzL0Sfz0XbcuSBnub"
upstream_id
"chatcmpl-BMJ5yQ3uH2yypqah6wXSGGoLeF51F"
total_cost
0.002726
cache_discount
(null)
provider_name
"OpenAI"
created_at
"2025-04-14T18:45:24.799712+00:00"
Monday, April 14th, 2025 at 6:45:24 PM GMT+00:00
model
"openai/gpt-4.1-2025-04-14"
app_id
182717
streamed
true
cancelled
false
latency
562
moderation_latency
79
generation_time
5472
tokens_prompt
68
tokens_completion
323
native_tokens_prompt
67
native_tokens_completion
324
native_tokens_reasoning
0
num_media_prompt
(null)
num_media_completion
(null)
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.002726
Evaluation details
Result
Evaluator
Details
Meta Data
98.7124%
Matches word count
n/a
neededClean
false
words
206
36.7879%
Dialogue to Total Word Ratio
Ratio: 100.00%, Deviation: 10.00%
neededClean
false
wordsTotal
206
wordsDialogue
206
67.7502%