NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-18-01KM07VHQAHKAV220S91ENGZG4
Run Details
Model:
GPT-5.4 Mini
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
2.6s
Cost:
$0.001507
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
9.61%
Matches word count
(details)
91.04%
Dialogue to Total Word Ratio
(details)
50.3225%
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, tell me you brought the map,” Molly whispered. “I brought the map, the snacks, and three different excuses,” Bill whispered back. “I am a professional.” “Professionals don’t wear clown shoes.” “These are stealth clown shoes.” “Those are clown shoes with the bells cut off.” “Exactly. Silent mode.” “Bill, the party’s in the church hall. There are balloons, streamers, and at least one inflatable unicorn.” “I hate unicorns.” “You hate everything that reminds you of joy.” “That’s because joy is suspicious.” “Focus. We slip in during the magic show, mingle with the parents, and move the presents into the van.” “Easy,” Bill said. “You distract them. I’ll do the lifting.” “With what strength?” “My emotional commitment.” “That’s not a thing.” “It is if you say it confidently.” “Mmm. Fine. But no mistakes. Last time you got caught by a six-year-old with a glitter beard.” “He was very agile.” “He was four.” “He had range.” “Bill.” “Fine. We take the presents, we leave the wrapping paper, we vanish.” “Like ghosts.” “Like responsible ghosts.” “Bill?” “Yeah?” “If you touch the cake, I’ll bury you in the sandbox.” “Noted. Though, for the record, I do think the cake is part of the loot.” “Bill.” “Just saying, Molly. A heist is a heist.” “And this one,” Molly said, grinning, “is going to be wrapped up beautifully.”
created_at
"2026-03-18T10:28:56.186Z"
3/18/2026, 10:28:56 AM
model
"openai/gpt-5.4-mini-20260317"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
859
moderation_latency
(null)
generation_time
2385
tokens_prompt
77
tokens_completion
440
native_tokens_prompt
66
native_tokens_completion
324
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
0.0015075
router
(null)
provider_responses
0
endpoint_id
"9ee065b2-3d1c-43bc-bdd7-28af3b148282"
id
"resp_02f01d754e21cdb20169ba7e6842ec8190a67695185b2539bc"
is_byok
false
latency
220
model_permaslug
"openai/gpt-5.4-mini-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-1773829736-8SPnadLv4jM2XACNoV1c"
upstream_id
"resp_02f01d754e21cdb20169ba7e6842ec8190a67695185b2539bc"
total_cost
0.0015075
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
9.61%
Matches word count
n/a
neededClean
false
words
222
91.04%
Dialogue to Total Word Ratio
Ratio: 95.54%, Deviation: 5.54%
neededClean
false
wordsTotal
224
wordsDialogue
214
50.3225%