NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1H0XRRHAAH2DE3HKZWXFHY
Run Details
Model:
GPT-5.4
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
11.3s
Cost:
$0.004830
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
81.27%
Matches word count
(details)
22.73%
Dialogue to Total Word Ratio
(details)
51.9992%
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 peered through the bakery window at the riot of balloons next door, where a clown stumbled past a cardboard castle and three mothers argued over napkins. She adjusted her black gloves with priestly solemnity. “Bill, this is our Sistine Chapel.” “It’s a village hall full of juice boxes.” “A treasure vault wrapped in cartoon paper,” Molly whispered. “Observe the terrain.” Bill squinted. “Bouncy castle at six o’clock. Sugar-crazed minors everywhere. One magician, possibly armed.” “We go in disguised as entertainers.” “You can’t juggle.” “I can lie confidently while dropping things.” Bill tugged his cap lower. “What about the presents?” “Table by the cake. We create a diversion, seize the loot, and vanish before the candles are lit.” “That’s your plan? Chaos and speed?” “That is every good plan.” Inside, a shriek rose as a child walloped the clown with a foam sword. Bill winced. “Hostile security.” Molly grinned. “Amateur security. Sticky, loud, and easily bribed with sherbet.” “With what money?” “We use their money. Check the birthday cards.” Bill stared. “That’s despicable.” “That,” Molly said, straightening her jacket, “is layered criminal thinking.” A woman emerged carrying a tray of cupcakes. Bill inhaled. “New plan. We steal one present and six cakes.” Molly considered it gravely. “At last, the voice of professionalism.”
created_at
"2026-03-06T12:12:42.143Z"
3/6/2026, 12:12:42 PM
model
"openai/gpt-5.4-20260305"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
422
moderation_latency
(null)
generation_time
11209
tokens_prompt
77
tokens_completion
410
native_tokens_prompt
66
native_tokens_completion
311
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
0.00483
router
(null)
provider_responses
0
endpoint_id
"9ff5625c-403f-4d7f-b895-58ac7295062c"
id
"resp_0f0010bf2a78968d0169aac4ba35a88195b3cf7c6ffac44ae6"
is_byok
false
latency
171
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-1772799162-fzOs9GSkCjFN5LO4729q"
upstream_id
"resp_0f0010bf2a78968d0169aac4ba35a88195b3cf7c6ffac44ae6"
total_cost
0.00483
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
81.27%
Matches word count
n/a
neededClean
false
words
212
22.73%
Dialogue to Total Word Ratio
Ratio: 61.03%, Deviation: 11.03%
neededClean
false
wordsTotal
213
wordsDialogue
130
51.9992%