NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHK98BJCB6RYNRMK3WES0SB2
Run Details
Model:
GPT-5.2
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
29.7s
Cost:
$0.023901
Provider:
OpenAI
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
They crouched behind the bouncy castle, listening to squeals and kazoo music. "Molly, why are we robbing eight-year-olds?" "Because, Bill, the loot is pre-wrapped and labeled. Efficiency." "Still feels wrong." "Wrong is leaving unclaimed LEGO sets on a table." "What's the entry plan?" "We go as entertainers. You, magician; me, balloon artist." "I can’t do magic." "You can vanish. You do it emotionally every time I ask for rent." "Harsh. Okay, disguise?" "Top hats, glitter vests, fake smiles. The parents will applaud anything." "And the presents?" "Phase two: 'Surprise Treasure Hunt.' We lead the kids outside." "With what, candy?" "With a dramatic map and your tragic card tricks." "While they're hunting, we scoop the gifts?" "Exactly. Bag them, walk out like heroes." "What about the birthday kid?" "I'll give them a balloon dog. Very therapeutic." "And if a parent asks?" "Say it's for 'game redistribution.' Use a serious voice." "Serious voice?" "'Ma’am, we’re optimizing fun.'" "Security?" "Clown hired from Craigslist. I checked his reviews: 'Mostly sober.'" "Exit strategy?" "Drive away in the ice-cream truck." "We don’t have an ice-cream truck." "We will in ten minutes, Bill. Keep up." "I'm sweating." "Great. Glitter sticks better." "Also, Bill, stop eating the piñata."
created_at
"2026-02-16T13:12:23.484418+00:00"
2/16/2026, 1:12:23 PM
model
"openai/gpt-5.2-20251211"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
26585
moderation_latency
(null)
generation_time
29637
tokens_prompt
77
tokens_completion
351
native_tokens_prompt
66
native_tokens_completion
1699
native_tokens_completion_images
0
native_tokens_reasoning
1378
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"completed"
usage
0.0239015
router
(null)
provider_responses
0
id
"resp_010e0d079eb7983e01699317993f688193a6951640ba94fdcb"
status
200
is_byok
false
latency
148
endpoint_id
"f00142c2-6a93-49ce-9e36-5593b904ce3b"
provider_name
"OpenAI"
model_permaslug
"openai/gpt-5.2-20251211"
api_type
"completions"
id
"gen-1771247513-o7LBzPyTrBqYALoFoiNX"
upstream_id
"resp_010e0d079eb7983e01699317993f688193a6951640ba94fdcb"
total_cost
0.0239015
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
99.7443%
Matches word count
n/a
neededClean
false
words
204
97.1662%
Dialogue to Total Word Ratio
Ratio: 94.12%, Deviation: 4.12%
neededClean
false
wordsTotal
204
wordsDialogue
192
98.4553%