NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1HV3B8SPB79YJDCCW8THR1
Run Details
Model:
GPT-5.4
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
10.4s
Cost:
$0.005070
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
65.70%
Dialogue to Total Word Ratio
(details)
32.8486%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 50% 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 crouched behind the hedge, studying the suburban garden where balloons bobbed like surveillance blimps and a clown shaped suspiciously like despair folded dogs for shrieking children. Bill adjusted his flat cap and peered over her shoulder at the mountain of brightly wrapped presents stacked beneath a banner screaming HAPPY SEVENTH, OLIVER. “Tell me again,” Bill whispered, “why we’re robbing a child instead of someone with a yacht?” “Because,” Molly said, “people with yachts hire guards. People with bouncy castles hire magicians.” “That clown looks mean.” “That clown looks unemployed.” A paper hat sailed past on the breeze and landed on Bill’s nose. He frowned at it like it had insulted his ancestry. “I still say the jelly table is a tactical nightmare.” “We avoid the jelly table.” “And the pony?” “We charm the pony.” Bill watched a herd of children stampede after bubbles. “You ever feel morally complicated?” “Constantly,” said Molly. “That’s why I dress well.” From the patio came the mother’s voice, distracted and desperate, counting juice boxes like a general losing a war. “Signal?” Bill asked. Molly nodded toward the chaos. “When the magician sets his sleeve on fire.” “That specific?” “He’s sweating confetti. It’s inevitable.” Right on cue, a child screamed with delight, smoke puffed upward, and parents surged toward the patio. “Showtime,” Molly said. Bill grinned. “For the presents, and possibly the pony.” “Not the pony.” “We’ll discuss it while running.”
created_at
"2026-03-06T12:26:59.824Z"
3/6/2026, 12:26:59 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
414
moderation_latency
(null)
generation_time
10343
tokens_prompt
77
tokens_completion
446
native_tokens_prompt
66
native_tokens_completion
327
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
0.00507
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_0e0cf2a9d9a645150169aac813e1c88196bfbd33a60ffa8ba8"
is_byok
false
latency
145
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-1772800019-jgGCrDLa1Huxno6zJPUV"
upstream_id
"resp_0e0cf2a9d9a645150169aac813e1c88196bfbd33a60ffa8ba8"
total_cost
0.00507
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
236
65.70%
Dialogue to Total Word Ratio
Ratio: 41.95%, Deviation: 8.05%
neededClean
false
wordsTotal
236
wordsDialogue
99
32.8486%