NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHKA29AQ67T0CEPBJR5KVG7V
Run Details
Model:
GPT-5
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
1m 25s
Cost:
$0.042412
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
They crouched behind a bouncy castle. "Molly, remind me why we’re robbing a birthday instead of a bank?" "Because banks have guards, alarms, and vaults; kids have cake, chaos, and adults negotiating juice-box treaties." "Bill, picture it: we stroll in, applaud the magician, and exit with a gift pyramid worthy of pharaohs." "Stroll? I trip over confetti at a funeral; I’ll need wheels." "We’ll borrow an inflatable unicorn labeled Emergency Parade; nobody questions a parade." "Clowns are risky; we’ll go as Concerned Aunts, fiercely punctual and armed with thank-you notes." "I can aunt-nod magnificently, but I can’t high-five a piñata without flinching." "Avoid the piñata; it’s a candy volcano with union rules." "What about the cake?" "Decoy frosting: I’ll compliment the layers until time itself stops to blush." "We assemble them into a humanitarian convoy bound for the Museum of Unopened Wonders." "We’ll leave a thank-you card and a voucher for replacement bows, morally neutral and extremely shiny." "Right after musical chairs; society dissolves, chairs betray, and nobody notices a wagon full of generosity." "What if a toddler tackles me?" "Compliment their shoes and retreat in zigzags; toddlers respect geometry." "You brilliant villain, save me a slice of plausible deniability."
created_at
"2026-02-16T13:27:28.005315+00:00"
2/16/2026, 1:27:28 PM
model
"openai/gpt-5-2025-08-07"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
26465
moderation_latency
(null)
generation_time
84452
tokens_prompt
77
tokens_completion
1227
native_tokens_prompt
66
native_tokens_completion
4233
native_tokens_completion_images
0
native_tokens_reasoning
3904
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
0.0424125
router
(null)
provider_responses
0
id
"resp_05104e550779933e0169931aeaf0708194849a5696dc35fc53"
status
200
is_byok
false
latency
204
endpoint_id
"7c2f859a-7890-4e8e-b1de-1cd1c0a800b4"
provider_name
"OpenAI"
model_permaslug
"openai/gpt-5-2025-08-07"
api_type
"completions"
id
"gen-1771248362-3eX1Cgpx9aASk3XxlnlK"
upstream_id
"resp_05104e550779933e0169931aeaf0708194849a5696dc35fc53"
total_cost
0.0424125
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
99.3769%
Matches word count
n/a
neededClean
false
words
205
77.8569%
Dialogue to Total Word Ratio
Ratio: 97.07%, Deviation: 7.07%
neededClean
false
wordsTotal
205
wordsDialogue
199
88.6169%