NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1GMCD0XRYRBH334PNS93CY
Run Details
Model:
GPT-5.4 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
10.5s
Cost:
$0.005370
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
51.93%
Matches word count
(details)
36.79%
Dialogue to Total Word Ratio
(details)
44.3567%
Overall
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, stop jingling. You sound like a tambourine with guilt.” “I can’t help it, Molly. These lock picks have ambition.” “We’re not robbing a bank. We’re raiding a pirate-themed sixth birthday.” “Exactly. Chaos, sugar, weak perimeter.” “We need finesse. Children can smell fear and off-brand aftershave.” “This is premium aftershave.” “It smells like a haunted lemon.” “Fine. Plan review. We enter disguised as entertainers.” “No. You juggle like a man fighting invisible bees.” “I’ve improved.” “You set your sleeve on fire.” “That was one time. Mascots, then?” “You’re six foot three. What mascot were you thinking, a probationary giraffe?” “Right. Then caterers.” “Better. Hairnets, trays, confident disappointment.” “I do confident disappointment naturally.” “I know. Now, presents are stacked by the fireplace, under the banner saying, ‘HAPPY BIRTHDAY, OLLIE, YOU LEGEND.’” “Bit much for six.” “Focus. We wait for cake. During the candle song, adults cry, children scream, nobody counts boxes.” “And we sweep the gifts into the trolley?” “The one labeled ‘juice boxes.’” “Elegant.” “If confronted, you say, ‘Allergy inspection.’” “To presents?” “Parents believe anything if you look exhausted.” “What if the birthday child spots us?” “Then improvise.” “With what?” “Balloon animal. Bribery. Empathy. Your sad lemon face.” “Molly?” “Yes?” “Are we the villains?” “Bill, we’re stealing from a bouncy castle economy. Try to have perspective.”
created_at
"2026-03-06T12:05:51.142Z"
3/6/2026, 12:05:51 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
456
moderation_latency
(null)
generation_time
10449
tokens_prompt
77
tokens_completion
476
native_tokens_prompt
66
native_tokens_completion
347
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
0.00537
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_08a2226a2f57ba120169aac31f3648819690ea4b823c266bb1"
is_byok
false
latency
162
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-1772798751-O58j056oKq6KoZ36FvJ9"
upstream_id
"resp_08a2226a2f57ba120169aac31f3648819690ea4b823c266bb1"
total_cost
0.00537
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
51.93%
Matches word count
n/a
neededClean
false
words
216
36.79%
Dialogue to Total Word Ratio
Ratio: 100.00%, Deviation: 10.00%
neededClean
false
wordsTotal
218
wordsDialogue
218
44.3567%