NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-18-01KM099HHMQPD88CBSPHJXYRTG
Run Details
Model:
GPT-5.4 Nano
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
2.7s
Cost:
$0.000649
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
8.92%
Dialogue to Total Word Ratio
(details)
4.4604%
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, confirm the plan: we don’t touch the cake,” Bill whispered. “Of course we don’t touch the cake. The cake is bait. The real prizes are in the present pile,” Molly said. “Agreed. We’re stealing all the presents, not learning any lessons,” Bill replied. “Technically we’re learning a lesson: never invite us again,” Molly said. “Great, because last time the clown gave me a balloon and I accidentally apologized to it,” Bill groaned. “That’s why we’re wearing disguises. Your disguise is ‘harmless dad,’ and mine is ‘suspicious raccoon,’” Molly said. “I don’t look harmless,” Bill said. “You look like you stole a sandwich from time itself,” Molly replied. “Okay, silent steps, soft breath. No singing. You always sing,” Bill warned. “I sing under my breath,” Molly corrected. “Under your breath is still singing,” Bill said. “Fine. I will be a criminal statue,” Molly declared. “You’re already shaped like one,” Bill said. “The kid’s party will be loud,” Molly murmured. “Exactly. We use the chaos as cover. I’ll grab the big ones,” Bill said. “And I’ll take the small ones,” Molly answered. “What if there’s a watchdog?” Bill asked. “There’s a dog? I thought it was just anxiety with fur,” Molly said. “Then we distract it with the squeaky toy,” Bill said. “We don’t squeak the toy,” Molly warned. “We quietly squeak it,” Bill corrected. “That’s not quietly,” Molly said. “It’s… politely squeaky,” Bill insisted. “Just don’t set off the gift wrap alarms,” Molly said. “Those alarms are imaginary,” Bill replied. “So is your confidence,” Molly shot back. “Listen, when the music starts, you slip in. I’ll act normal,” Bill said. “You can’t act normal,” Molly replied. “I can act mildly suspicious,” Bill said. “Perfect. Now—on three,” Molly said. “Three,” Bill echoed. “One,” Molly continued. “Stop counting wrong,” Bill hissed. “Too late. We’re already stealing,” Molly said. “Alright. Stealing… gently… like burglars with manners,” Bill whispered.
created_at
"2026-03-18T10:54:03.328Z"
3/18/2026, 10:54:03 AM
model
"openai/gpt-5.4-nano-20260317"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
471
moderation_latency
(null)
generation_time
2320
tokens_prompt
77
tokens_completion
605
native_tokens_prompt
66
native_tokens_completion
509
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
0.00064945
router
(null)
provider_responses
0
endpoint_id
"0c835f2e-c18d-4e8c-b245-e1e3bd08b97f"
id
"resp_0c093a6118c1ca110169ba844b6d48819388003b0bf9ee0698"
is_byok
false
latency
186
model_permaslug
"openai/gpt-5.4-nano-20260317"
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-1773831243-qX2I3dnq1VVkBkmTV8yr"
upstream_id
"resp_0c093a6118c1ca110169ba844b6d48819388003b0bf9ee0698"
total_cost
0.00064945
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
316
8.92%
Dialogue to Total Word Ratio
Ratio: 77.53%, Deviation: 12.47%
neededClean
false
wordsTotal
316
wordsDialogue
245
4.4604%