NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1J04YCQ4YRPQ83AV1YHJA1
Run Details
Model:
GPT-5.4 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
9.9s
Cost:
$0.004935
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
43.38%
Matches word count
(details)
36.79%
Dialogue to Total Word Ratio
(details)
40.0831%
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, this is our classiest job yet.” “Classy? Molly, we’re robbing a bouncy castle.” “An affluent bouncy castle. Look at the ribbon budget.” “I still say stealing from a children’s party sounds cursed.” “We’re not stealing from children. We’re liberating gifts from inattentive adults.” “That sentence wore a monocle.” “Focus. We go in as entertainers.” “I can juggle three oranges and a grievance.” “Perfect. I’ll be the face painter.” “You can’t paint.” “I can paint panic. Same wrist action.” “What about the dog?” “What dog?” “The enormous poodle in a bow tie guarding the present table.” “That’s not a guard dog, Bill. That’s a soufflé with opinions.” “It just looked at me like I owe it rent.” “Then avoid eye contact. Rule one of burglary and exes.” “How do we carry everything out?” “Gift sacks. We swap the birthday banner for one reading ‘Prize Collection Point.’” “That’s diabolical.” “Thank you.” “What if a kid asks where their present’s going?” “You say, ‘Quality control.’ Children respect bureaucracy.” “I don’t think they do.” “They fear clipboards. Everyone fears clipboards.” “And if the clown recognizes us?” “We recognize him first. Establish dominance.” “With what, larger shoes?” “With confidence, Bill.” “I once tripped over confidence.” “I know. I was there. Now straighten your tie.” “It squirts water.” “Good. Professionalism and hydration.”
created_at
"2026-03-06T12:29:45.301Z"
3/6/2026, 12:29:45 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
412
moderation_latency
(null)
generation_time
9797
tokens_prompt
77
tokens_completion
460
native_tokens_prompt
66
native_tokens_completion
318
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
0.004935
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_09aa9753332629600169aac8b95dec81958b040b697dc56699"
is_byok
false
latency
176
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-1772800185-J3DmBfybeqkb6qIqML2X"
upstream_id
"resp_09aa9753332629600169aac8b95dec81958b040b697dc56699"
total_cost
0.004935
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
43.38%
Matches word count
n/a
neededClean
false
words
217
36.79%
Dialogue to Total Word Ratio
Ratio: 100.00%, Deviation: 10.00%
neededClean
false
wordsTotal
217
wordsDialogue
217
40.0831%