NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37HF82VFKAGZ8YNYJKN5X8S
Run Details
Model:
GPT-6 Sol
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
13.9s
Cost:
$0.006722
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
20.18%
Dialogue to Total Word Ratio
(details)
60.0919%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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 spread a hand-drawn map across the diner table. A purple crayon circle marked the living room. A yellow star marked the cake. Around the edges, in green, she had written: BEWARE OF GLITTER. Bill studied it over his pancakes. “You’ve been inside?” “I delivered balloons. The birthday girl asked if I was a princess.” “What did you say?” “I told her princesses don’t carry helium tanks.” “Cold.” “She tipped me a sticker.” Molly tapped a blue rectangle. “Presents go here, beside the sofa. The party starts at two. We arrive at two fifteen, when the parents are busy opening juice boxes.” Bill lowered his fork. “You’re planning to steal from children.” “I’m planning to steal wrapped presents. Their owners can be anyone until the tags are read.” “That sounds like something a villain says before becoming a frog.” Molly ignored him. Across the diner, a child’s balloon escaped a booth and drifted to the ceiling. Bill watched it rise with the expression he usually reserved for unpaid parking tickets. “Costumes,” Molly said. “We go as entertainers.” “You can juggle?” “I can face-paint.” “You painted a tiger on my face once. People thought I had a rash.” “The lighting was poor. You’ll be the magician.” Bill looked at his hands. “I know one trick.” “The disappearing coin?” “No. I can pull a tablecloth off a table without breaking anything, provided there’s nothing on the table.” Molly drew a second arrow on the map, leading from the living room to the back gate. She labeled it EXIT, then crossed it out and wrote GRAND FINALE. Their previous plans had suffered from insufficient showmanship, according to Molly, and excessive police attention, according to Bill. At two fifteen, they stood on the porch of a house shaped like a cupcake that had grown doors. Bill wore a cape made from a shower curtain. Molly carried a makeup case and a sack large enough to suggest troubling ambitions. The door opened. A woman with icing on her elbow smiled. “Oh, thank goodness. You’re early.” Molly blinked. “Early?” “The other magician canceled. Come in!” Children swarmed Bill before he could object. One handed him a wand. Another demanded a rabbit. Molly found herself seated beneath a paper banner, painting whiskers on a boy who insisted he was a shark. “Sharks don’t have whiskers,” she whispered. “This one does,” he said. For forty minutes, Bill made coins vanish by dropping them into his cape pocket. The children cheered every time. Molly painted six tigers, three butterflies, and a shark with whiskers. Then came presents. Bill caught Molly’s eye across the room. The sack waited by the door, empty. The birthday girl unwrapped a plastic kitchen set. “You can have the chef hat,” she told Molly. Molly put it on. Bill’s wand slipped from his fingers. Outside, once they’d been paid, he said, “We still stole something.” “What?” Bill held up a juice box. Molly snatched it, stuck in the straw, and drank as they walked away.
created_at
"2026-09-23T16:27:07.238Z"
9/23/2026, 4:27:07 PM
model
"openai/gpt-6-sol-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2396
moderation_latency
(null)
generation_time
13840
tokens_prompt
75
tokens_completion
810
native_tokens_prompt
66
native_tokens_completion
659
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)
num_fetches
(null)
web_search_engine
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"completed"
service_tier
"default"
usage
0.006722
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790180827-7VhxiNsIZudHRUViOESo"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790180827-6LclonZzPvPxvKDKUrHm"
upstream_id
"resp_024aed28b16fb324016ab3fddb5dc487d19899f45f9dacd7c1"
provider_responses
0
endpoint_id
"1cfc7d9d-4404-4b8e-9ee4-58ae45c9dcd4"
id
"resp_024aed28b16fb324016ab3fddb5dc487d19899f45f9dacd7c1"
is_byok
false
latency
665
model_permaslug
"openai/gpt-6-sol-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.006722
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
20.18%
Dialogue to Total Word Ratio
Ratio: 35.66%, Deviation: 5.66%
neededClean
false
wordsTotal
502
wordsDialogue
179
60.0919%