NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1HBBPWH22TBXJSJBKF1X00
Run Details
Model:
GPT-5.4 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
8.4s
Cost:
$0.004440
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
14.30%
Matches word count
(details)
74.89%
Dialogue to Total Word Ratio
(details)
44.5937%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 and Bill crouched behind the village hall hedge, dressed in black so severe they looked like undertakers for a magician. Through the window came shrieks, balloons, and the unmistakable chaos of six-year-olds discovering sugar. A papier-mâché dragon guarded a table piled high with brightly wrapped presents, each one glowing in Bill’s imagination like treasure in a pirate’s fever dream. Molly unfolded the hand-drawn plan with the solemnity of a general invading a very small, very sticky nation. The back gate led to the kitchen, the kitchen led to the hall, and the hall led to glory, assuming they avoided the clown, who seemed to move with the twitchy vigilance of a prison warden. “Grab and go,” Molly whispered. Bill adjusted his wool cap and peered at the window. “I hate clowns.” Inside, a woman in a tiara was wrestling a cake away from a child with both fists in the icing. Near the presents, a boy in a cape patrolled in circles, brandishing a plastic sword with the grim authority of a customs officer. Molly narrowed her eyes. Timing was everything. Bill, unfortunately, was Bill. As he rose to sneak toward the gate, a balloon under his knee exploded like a pistol shot. Twenty children turned at once. The clown smiled. Molly sighed, already calculating the cost of abandoning him.
created_at
"2026-03-06T12:18:24.099Z"
3/6/2026, 12:18:24 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
426
moderation_latency
(null)
generation_time
8387
tokens_prompt
77
tokens_completion
341
native_tokens_prompt
66
native_tokens_completion
285
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
0.00444
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_0242086d7960c2860169aac61029a081939c08b3ed6f881e26"
is_byok
false
latency
147
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-1772799504-ItQPIjsPcUE6d35hcBl6"
upstream_id
"resp_0242086d7960c2860169aac61029a081939c08b3ed6f881e26"
total_cost
0.00444
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
14.30%
Matches word count
n/a
neededClean
false
words
221
74.89%
Dialogue to Total Word Ratio
Ratio: 2.67%, Deviation: 7.33%
neededClean
false
wordsTotal
225
wordsDialogue
6
44.5937%