NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-09-01KKABY7K1T02XKX1WK8NJECJB
Run Details
Model:
GPT-5.4 (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
18.7s
Cost:
$0.018780
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.43%
Dialogue to Total Word Ratio
(details)
99.7129%
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
“Molly, casing a toy store is one thing, but robbing a six-year-old’s birthday?” “Seven-year-old,” said Molly. “Respect the mark.” “That makes it worse.” “It makes it organized. Balloons at noon, cake at one, sugar crash at two. We strike at two-oh-five.” “With what, conscience suppressors?” “With clown wigs. No one suspects extra clowns.” “Everyone suspects extra clowns.” “Fine. Magicians.” “I can’t do card tricks.” “You don’t need tricks, Bill. You need a van and the ability to say, ‘These gifts require immediate quality testing.’” “To children?” “To parents. Children can smell weakness.” Bill peered at the invitation. “Treasure Hunt, bouncy castle, pony rides. This kid’s birthday has a larger budget than my last burglary.” “Exactly. Rich parents, expensive presents, distracted guests. It’s practically grant-funded crime.” “And the escape?” “Through the garden, over the hedge, into the alley.” “The pony rides concern me.” “Then avoid the ponies.” “They never avoid me.” Molly grinned and folded the map. “Bill, tonight we steal teddy bears, remote-control dinosaurs, and possibly a karaoke machine.” Bill sighed. “If I’m trampled by a Shetland, put ‘died professionally’ on my grave.” “And if we’re caught?” “We say we’re entertainers.” “For whom?” “For the presents, Bill. They’re leaving with us.”
created_at
"2026-03-09T22:37:00.906Z"
3/9/2026, 10:37:00 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
16622
moderation_latency
(null)
generation_time
18552
tokens_prompt
77
tokens_completion
420
native_tokens_prompt
66
native_tokens_completion
1241
native_tokens_completion_images
0
native_tokens_reasoning
928
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
0.01878
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_006ef5824eba5d440169af4b8cf7fc8196a40aef6a73a6adf3"
is_byok
false
latency
153
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-1773095820-bTnzGE1Y8xqBwNCoq5Tl"
upstream_id
"resp_006ef5824eba5d440169af4b8cf7fc8196a40aef6a73a6adf3"
total_cost
0.01878
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
201
99.43%
Dialogue to Total Word Ratio
Ratio: 92.75%, Deviation: 2.75%
neededClean
false
wordsTotal
207
wordsDialogue
192
99.7129%