NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-09-01KK9QBE74WTBCNAR9N6YHPVF7
Run Details
Model:
GPT-5.4 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
25.4s
Cost:
$0.023265
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
99.98%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
99.9920%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 50% 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
Outside the church hall, balloons bobbed like surveillance blimps while a rented clown wrestled a folding sign that read HAPPY SEVENTH, OLIVER. Molly crouched behind the hedge, notebook open, wearing a caterer’s apron over burglar black. Bill balanced beside her with a coil of ribbon, a fake moustache, and the solemn expression of a man about to betray every child in Hampshire. Through the steamed windows came shrieks, party whistles, and the thump of a hired magician bombing spectacularly. On the gift table inside, parcels glittered in a mound of bright paper, each bow winking like an alarm begging ignored tonight. “Molly, this feels morally sticky.” “Bill, they’re children; they’ll forget by Tuesday.” “They remember ponies forever.” “Focus. We enter as entertainers, dazzle the adults, lift the loot, vanish before cake.” “I can juggle oranges, not suspicion.” “Then juggle both.” “What if the birthday boy cries?” “We leave him the batteries.” “That’s monstrous.” “That’s thoughtful.” “I hate that you make crime sound organized.” “I hate that you make panic sound musical.” “Fine. What’s my role?” “You’re Professor Sprinkles, balloon animal savant.” “And you?” “Captain Cupcake, widow, mysterious, excellent with knots.” “Why widow?” “Respectability.” “Right. Pillage, then politeness.” “Exactly. Smile big, steal bigger.”
created_at
"2026-03-09T16:37:13.583Z"
3/9/2026, 4:37:13 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
22976
moderation_latency
(null)
generation_time
25210
tokens_prompt
77
tokens_completion
522
native_tokens_prompt
66
native_tokens_completion
1540
native_tokens_completion_images
0
native_tokens_reasoning
1238
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
0.023265
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_08d0faf4daf829e90169aef739b5c481969cadd341263f2172"
is_byok
false
latency
204
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-1773074233-fFFXNjcgm2KoZcu1UKUj"
upstream_id
"resp_08d0faf4daf829e90169aef739b5c481969cadd341263f2172"
total_cost
0.023265
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
99.98%
Matches word count
n/a
neededClean
false
words
202
100.00%
Dialogue to Total Word Ratio
Ratio: 50.00%, Deviation: 0.00%
neededClean
false
wordsTotal
202
wordsDialogue
101
99.9920%