NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37J2YH3RG61YNS5HY12RR26
Run Details
Model:
GPT-6 Sol (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
19.3s
Cost:
$0.009082
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
5.88%
Dialogue to Total Word Ratio
(details)
2.9401%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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 crayon map across the café table. Bill moved his coffee off the swimming pool. “That’s not a swimming pool,” Molly said. “That’s the ball pit.” “Why is it blue?” “Because I only had three crayons. Focus. Saturday, two o’clock. We go in, we take the presents, we leave.” “All of them?” “All of them.” Bill studied the map. “There are seventeen children.” “Sixteen. The one with the crown is a cardboard princess.” “You’ve been inside?” “I attended the reconnaissance party.” “You went to a child’s birthday party without me?” “I told you there’d be cake.” “You said the cake was a lie.” “It was. They had cupcakes.” Bill pushed the map away. Outside, rain slid down the café window, making the bakery opposite look as if it were melting. “I don’t like this,” he said. “Stealing from children.” “We’re stealing from their parents. Children don’t have money.” “They’ll notice their presents are gone.” “Only briefly. They have the attention spans of sparrows.” “My niece remembers the exact location of a biscuit I ate in 2021.” Molly lowered her voice. “These aren’t birthday presents. The parents bring one wrapped gift each for a raffle. The proceeds go to the school roof.” “That makes it worse.” “How?” “The roof’s for children.” “The gifts are donated by the parents, who can afford them. We sell the gifts, then pay for the roof ourselves. We keep whatever’s left.” Bill stared at her. “You want to steal a raffle to donate to its own cause?” “Yes.” “Why not just buy tickets?” “Because my mother’s running it.” “That explains the wig.” Molly touched her enormous blond curls. “She’d recognise me.” “She’d recognise you from space. You’re wearing her earrings.” “They complete the disguise.” “They have your name on them.” Molly tucked the earrings into her coat. “Fine. New problem: the presents are beside the buffet. Your job is distraction.” “What kind?” “Magic.” “I know one trick.” “It’s a good trick.” “You screamed when I did it.” “Because my watch disappeared.” “It came back.” “Three days later, in your shoe.” Bill leaned over the map. “What if I dress as the cardboard princess?” “The princess has no legs.” “I’ll stand behind the table.” “You’ll need to stand there for three hours.” “I’ve waited longer for a bus.” Molly considered this. Bill brightened. “No,” she said. “You distract the adults. I’ll carry the presents out.” “In what?” She produced a folded canvas bag. Bill unfolded it until it covered both their laps and brushed the floor. “That holds a bicycle.” “Exactly.” “A bicycle would be less suspicious.” “Then put a bell on the bag.” Bill folded it again. “What happens when your mother catches us?” “She won’t.” “And if she does?” Molly looked out at the rain. “She’ll make us stack chairs.” Bill shuddered. “Last year she made me judge the costume contest.” “You gave every child first prize.” “They all looked disappointed.” “Because you ate the prizes.” “They were chocolate medals. I thought they were for the judges.” Molly rolled up the map. “Two o’clock. Bring the watch.” “For the trick?” “No. So we know when to run.” Bill checked his wrist. “It’s still in my other shoe.”
created_at
"2026-09-23T16:37:52.813Z"
9/23/2026, 4:37:52 PM
model
"openai/gpt-6-sol-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
5923
moderation_latency
(null)
generation_time
19298
tokens_prompt
75
tokens_completion
956
native_tokens_prompt
66
native_tokens_completion
895
native_tokens_completion_images
0
native_tokens_reasoning
170
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
0.009082
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790181472-3d2H0fOgNgB5DN3zsCcS"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790181472-bdFQB29QUDS0Os9bCS0t"
upstream_id
"resp_00f7db8df47914b5016ab40060f07c87d1aa5e189b6cc8bf97"
provider_responses
0
endpoint_id
"1cfc7d9d-4404-4b8e-9ee4-58ae45c9dcd4"
id
"resp_00f7db8df47914b5016ab40060f07c87d1aa5e189b6cc8bf97"
is_byok
false
latency
1216
model_permaslug
"openai/gpt-6-sol-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.009082
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
0.00%
Matches word count
n/a
neededClean
false
words
534
5.88%
Dialogue to Total Word Ratio
Ratio: 77.53%, Deviation: 7.53%
neededClean
false
wordsTotal
534
wordsDialogue
414
2.9401%