NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHKAW3KM3BTY8YTEC1BVPCZM
Run Details
Model:
GPT-5
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
56.5s
Cost:
$0.036382
Provider:
OpenAI
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
They staked out the inflatable castle from a bush that smelled like juice boxes and desperation. Molly sketched routes on a napkin sticky with frosting, labeling choke points: clown, cake, bewildered parents. Bill calibrated a grab-bag made from a laundry hamper and a dog leash. Balloons bobbed like buoyant sentries. A tiny wizard practiced vanishing; the trick box, crucially, was open. "Masks with unicorns?" Molly whispered. "Less glitter, more getaway," Bill replied. "We blend, grab, vanish," Molly said. "Confetti makes footprints," Bill warned. "Then moonwalk," Molly proposed. "Moonboots squeak," Bill countered. "Tiptoe," Molly said. "Agreed," Bill sighed. Molly produced a sack patterned with dinosaurs to appease the gods of camouflage. The plan: infiltrate during the birthday chant, when adults videotape and children scream in one jubilant key. Bill would fake a balloon emergency, tugging every ribbon so the ceiling swallowed evidence. Molly would slide under the gift table like a very polite ferret, convert presents to a tidy avalanche, and surf them to the back gate. For exit music, the DJ’s Baby Shark remix would cover the rattling. If questioned, Bill would point to a name tag reading Uncle Probably, while Molly juggled cupcakes into plausible credentials. What could misfire?
created_at
"2026-02-16T13:41:05.820786+00:00"
2/16/2026, 1:41:05 PM
model
"openai/gpt-5-2025-08-07"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3688
moderation_latency
(null)
generation_time
56271
tokens_prompt
77
tokens_completion
1140
native_tokens_prompt
66
native_tokens_completion
3630
native_tokens_completion_images
0
native_tokens_reasoning
3328
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
0.0363825
router
(null)
provider_responses
0
id
"resp_051a90797f0b01480169931e390eb48193b361baec43149839"
status
200
is_byok
false
latency
194
endpoint_id
"7c2f859a-7890-4e8e-b1de-1cd1c0a800b4"
provider_name
"OpenAI"
model_permaslug
"openai/gpt-5-2025-08-07"
api_type
"completions"
id
"gen-1771249208-YQgBBjq2FxE4AhzZrz7p"
upstream_id
"resp_051a90797f0b01480169931e390eb48193b361baec43149839"
total_cost
0.0363825
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
99.9990%
Matches word count
n/a
neededClean
false
words
201
100.0000%
Dialogue to Total Word Ratio
Ratio: 9.95%, Deviation: 0.05%
neededClean
false
wordsTotal
201
wordsDialogue
20
99.9995%