NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1GFXXACNCZVVJR3TXJ5MX5
Run Details
Model:
GPT-5.4 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
10.1s
Cost:
$0.004740
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
99.34%
Dialogue to Total Word Ratio
(details)
49.6679%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 and Bill crouched behind the florist’s van, staring at the suburban fortress ahead: a garden full of bunting, shrieking children, and a clown so cheerful he looked legally suspicious. Through the hedges they could see a mountain of brightly wrapped presents piled beside a plastic throne where the birthday boy sat wearing a paper crown and the expression of a tiny emperor. Molly unfolded a crayon-scribbled map stolen from the community noticeboard. Her eyes gleamed with professional greed. Bill adjusted his black wool cap, which made him look less like a master thief and more like a nervous chimney sweep at a funeral. “Too many balloons,” Bill muttered. “Balloons are not security,” Molly said. A magician wandered past the gate, pulling scarves from his sleeve while a terrier chased his ankles with religious fury. Near the patio, a woman in a ladybird costume arranged cupcakes with military concentration. Every route to the presents involved crossing open lawn, dodging sugar-hypnotized children, and possibly surviving the clown. Bill peered again. “What if they see us?” Molly smiled, producing two cone hats and a gift bag containing absolutely nothing. “Then we become disappointing relatives.” From inside the party came the crack of a piñata and the wild roar of small people discovering violence as a hobby. Molly took a breath, straightened her hat, and marched for the gate, with Bill wobbling behind her like regret in sensible shoes.
created_at
"2026-03-06T12:03:25.233Z"
3/6/2026, 12:03:25 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
417
moderation_latency
(null)
generation_time
10087
tokens_prompt
77
tokens_completion
386
native_tokens_prompt
66
native_tokens_completion
305
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)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"completed"
usage
0.00474
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_0261dc5261c607b60169aac28d4d348197a271b2ec661082d9"
is_byok
false
latency
179
model_permaslug
"openai/gpt-5.4-20260305"
provider_name
"OpenAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1772798605-RmAT4RpOHRGOtE7vUWvD"
upstream_id
"resp_0261dc5261c607b60169aac28d4d348197a271b2ec661082d9"
total_cost
0.00474
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
236
99.34%
Dialogue to Total Word Ratio
Ratio: 7.14%, Deviation: 2.86%
neededClean
false
wordsTotal
238
wordsDialogue
17
49.6679%