NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1H0N8TNYVC5TQ49HS2C8VS
Run Details
Model:
GPT-5.4 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
9.9s
Cost:
$0.004590
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.49%
Matches word count
(details)
90.54%
Dialogue to Total Word Ratio
(details)
45.5180%
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 crouched behind the church hall’s recycling bins, studying the birthday banner flapping over the gate like a taunt. Unicorns, balloons, a hired clown, and somewhere beyond the paper streamers, a mountain of brightly wrapped loot waiting beside a bouncy castle. Bill adjusted his wool cap and peered through the hedge with the solemn concentration of a man inspecting crown jewels rather than gift bags decorated with cartoon dinosaurs. “Quiet in, quick out,” Molly whispered. “Children are basically tiny drunks,” Bill murmured. “Unpredictable.” Inside, the party frothed with chaos. A magician was extracting scarves from his sleeve while three sugar-hysterical boys fenced with pool noodles. Near the trestle table sat the presents: glittering parcels stacked in a glorious heap, each one tagged in loopy handwriting, each one practically begging for a criminal misunderstanding. Molly straightened her cardigan, suddenly every inch the competent aunt. Bill carried a tray he’d pinched from the caterers and wore the strained smile of a substitute father at sports day. They slipped through the gate unnoticed, swallowed by shrieking games and the smell of cake icing. At the gift table Molly began loading parcels under the tray’s silver dome with breathtaking speed. Bill distracted a suspicious clown by asking directions to the toilet, then nodded sharply. In seconds they were gone, marching into the afternoon with enough presents to disappoint an entire postcode.
created_at
"2026-03-06T12:12:33.441Z"
3/6/2026, 12:12:33 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
406
moderation_latency
(null)
generation_time
9852
tokens_prompt
77
tokens_completion
385
native_tokens_prompt
66
native_tokens_completion
295
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
0.00459
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_029d81f6856d81f00169aac4b17e548194a60b81ed108ad6d6"
is_byok
false
latency
142
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-1772799153-dqpM5UIDHV45NMLI2SGs"
upstream_id
"resp_029d81f6856d81f00169aac4b17e548194a60b81ed108ad6d6"
total_cost
0.00459
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
0.49%
Matches word count
n/a
neededClean
false
words
227
90.54%
Dialogue to Total Word Ratio
Ratio: 4.39%, Deviation: 5.61%
neededClean
false
wordsTotal
228
wordsDialogue
10
45.5180%