NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHKAXKNSV79745VQE49V9YT2
Run Details
Model:
GPT-5 Nano
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
2m 5s
Cost:
$0.005383
Provider:
OpenAI
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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
The alley smells of popcorn and citrus, as their plan unfurls under a neon moon. Mirth dances in their eyes. Molly: "Bill, the giggling party bus is near." Bill: "Then we dash, like clowns crossing caution tape." Molly: "Clowns with manners, darling; timing is everything." Bill: "I brought whistle-safe glitter, and a ladder." Molly: "Glitter confuses guards; ladders confuse children." Bill: "We'll melt into the favors, like cookies." Molly: "No alarms, no alarms; just squeaky sneakers." Bill: "What about the cake confetti cannon?" Molly: "That is strictly a decoy; ignore it." Bill: "Decoys and decaf; we are caffeinated." Molly: "Remember the address: Grand Hall, not Grandma's." Bill: "Right, hall with candy chandelier." Molly: "Candy chandeliers never noticed shadows rehearsing bravery." Bill: "We practice, then exit before snacks tonight." Molly: "I will distract with a lullaby playlist." Bill: "You sing, I shimmy; crooners go unnoticed." Molly: "I'll booby-trap gift wrap, too." Bill: "Booby-trapped ribbons squeak on touch." Molly: "Should we warn the kids about impostor Santa?" Bill: "Only if he signs." Molly: "Fine, but keep the capes invisible." Bill: "Invisible capes are the oldest security trick." Molly: "Then we toast with juice boxes loudly." Bill: "Deal; we split glittery presents loot." Molly: "If anyone notices, we'll appeal to unicorns anyway." Bill: "Unicorns always take bribes of stickers." Molly: "Okay, we're done; party awaits." Bill: "Let's roll before cake collapses."
created_at
"2026-02-16T13:43:02.927176+00:00"
2/16/2026, 1:43:02 PM
model
"openai/gpt-5-nano-2025-08-07"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
11755
moderation_latency
(null)
generation_time
124150
tokens_prompt
77
tokens_completion
2401
native_tokens_prompt
66
native_tokens_completion
13449
native_tokens_completion_images
0
native_tokens_reasoning
12992
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
0.0053829
router
(null)
provider_responses
0
id
"resp_0bccfc0265f5e4530169931e6a3f7c8194ac97377276f6fa76"
status
200
is_byok
false
latency
120
endpoint_id
"50329d77-04e1-4979-a184-c33030289476"
provider_name
"OpenAI"
model_permaslug
"openai/gpt-5-nano-2025-08-07"
api_type
"completions"
id
"gen-1771249258-PTFcXgm2ceSy6PRNAXcY"
upstream_id
"resp_0bccfc0265f5e4530169931e6a3f7c8194ac97377276f6fa76"
total_cost
0.0053829
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
0.0098%
Matches word count
n/a
neededClean
false
words
231
25.9228%
Dialogue to Total Word Ratio
Ratio: 79.22%, Deviation: 10.78%
neededClean
false
wordsTotal
231
wordsDialogue
183
12.9663%