NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-09-01KKAC49B283JGNE134QHGAEYC
Run Details
Model:
GPT-5.4 (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
9.8s
Cost:
$0.009480
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
97.63%
Matches word count
(details)
99.66%
Dialogue to Total Word Ratio
(details)
98.6434%
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 crouched behind the bouncy castle like a burglar in a migraine, studying the suburban battlefield through binoculars decorated with cartoon sharks. Streamers whipped across the garden, a hired magician fought with a balloon octopus, and twenty sugar-crazed children ricocheted between folding chairs with the tactical discipline of startled pigeons. Beside her, Bill unfolded a hand-drawn map of the house that looked suspiciously like a treasure island designed by a tired pirate. “Front gate’s chaos,” Molly whispered. “We go through the caterer, charm the clown, grab the loot, vanish smiling.” Bill adjusted his clip-on bow tie, the kind criminals wore when they wanted to look honest and instead looked recently fired. The presents were stacked by the patio doors in a glittering mountain of dinosaurs, dolls, and educational plastic destined to be immediately ignored. A mechanical pony guarded them with dead-eyed serenity. Bill nodded, then frowned at the clown. The clown was tall, broad, and smoking behind the shed with the solemn air of a man reconsidering every decision since birth. Molly admired that. Real professionals respected despair. The gate burst open as another aunt arrived carrying cake. Bill straightened, inhaled courage, and marched toward destiny, trailed by Molly and an inflatable sword stuck to his shoe.
created_at
"2026-03-09T22:40:19.307Z"
3/9/2026, 10:40:19 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
8176
moderation_latency
(null)
generation_time
9435
tokens_prompt
77
tokens_completion
347
native_tokens_prompt
66
native_tokens_completion
621
native_tokens_completion_images
0
native_tokens_reasoning
349
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
0.00948
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_0fb7fd663ac838b20169af4c536130819588cb585d8766f679"
is_byok
false
latency
246
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-1773096019-fvg5pFt7scdohOcfTkP0"
upstream_id
"resp_0fb7fd663ac838b20169af4c536130819588cb585d8766f679"
total_cost
0.00948
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
97.63%
Matches word count
n/a
neededClean
false
words
207
99.66%
Dialogue to Total Word Ratio
Ratio: 7.58%, Deviation: 2.42%
neededClean
false
wordsTotal
211
wordsDialogue
16
98.6434%