NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX40BJ4280K0VPM5PNNWQT0P
Run Details
Model:
GPT-5.6 Sol
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
21.6s
Cost:
$0.031170
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0021%
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 the crayon map across the café table. “The party begins at two. Twelve children, eighteen adults, one magician, and approximately forty-seven presents.” Bill squinted at the drawing. “Why is the house purple?” “I borrowed the map from the birthday girl.” “You interviewed the target?” “She was feeding ducks.” “Did she suspect anything?” “She’s six, Bill. She suspected I was a mermaid.” “Are you?” “Focus.” “All right. We enter through the kitchen disguised as caterers.” “There are no caterers. It’s a budget party.” “Magician’s assistants?” “The magician works alone.” “Clowns?” “Absolutely not.” “Why?” “Clowns attract scrutiny.” “Clowns attract screaming. Screaming provides cover.” Molly tapped the map. “We go in as parents.” Bill stared at her. “Whose parents?” “Each other’s.” “That’s not how parents work.” “It is in several confusing families.” A waiter placed two coffees between them. Bill waited until he left, then whispered, “What are we stealing?” “All the presents.” “Yes, but what’s in them?” “Toys. Games. Possibly cash from lazy grandparents.” “Risky. Could be socks.” “Children don’t wrap socks in dinosaur paper.” “My aunt did.” “Your aunt sent you birthday socks?” “She said disappointment built character.” “It built you, certainly.” Bill examined the crayon map again. “Where are the presents?” “Dining room, beside the cake.” “The cake complicates matters.” “Because?” “I love cake.” “You cannot eat evidence.” “Cake isn’t evidence.” “Your moustache covered in blue icing would be.” “I can shave.” “Beforehand.” “That seems extreme.” Molly lowered her voice. “At two-fifteen, the magician starts. Everyone faces the garden. We slip into the dining room, load the presents into laundry sacks, and leave through the side door.” “Why laundry sacks?” “Festive efficiency.” “What if a child sees us?” “Tell them we’re Santa’s quality-control officers.” “In April?” “Santa plans ahead.” “Children ask questions.” “Then distract them.” “With what?” “Balloon animals.” “I can’t make balloon animals.” “You once made a giraffe.” “That was a plumbing accident.” Outside, a bus hissed to a stop. Molly folded the map, but Bill pinned one corner. “Security?” “One golden retriever.” “Bribable?” “Cheese-based morality.” “Excellent. Alarms?” “Motion sensor in the hall.” “How do we avoid it?” “We don’t. It’s a nightlight shaped like a moon.” “Then why mention it?” “To keep you alert.” “I’m extremely alert.” “You put sugar in the saltshaker.” “I was testing you.” “You salted your coffee.” “I was testing myself.” Molly retrieved a photograph. “This is the birthday boy’s father. Avoid him.” Bill studied it. “He looks familiar.” “He’s a police inspector.” Bill pushed back his chair. “New plan: we don’t rob the party.” “He’s off duty.” “He remains a police inspector recreationally.” “He’ll be watching the magician.” “Unless the magician is also a thief.” “He isn’t.” “How do you know?” “I checked the guild.” “There’s a thieves’ guild?” “Magicians’ guild.” “Less reassuring.” Molly slid him a second photograph. Bill groaned. “What now?” “The magician?” “Your brother.” “He owes me money.” “So this is personal.” “This is accounting.” “And the presents?” “Collateral whimsy.” Bill drained his salted coffee and shuddered. “Suppose we succeed. How do we divide the loot?” “Fifty-fifty.” “I want first choice.” “No.” “I need something for my nephew.” “You don’t have a nephew.” “I may acquire one.” “You cannot steal a child.” “I meant by marriage.” “You’re unmarried.” “I plan ahead. Like Santa.” Molly stood and buttoned her coat. “Meet me at the corner at one-fifty. Wear something parental.” “What does that mean?” “Cardigan. Tired eyes. Crumbs.” “I can manage crumbs.” “No weapons.” “Not even the umbrella pistol?” “Especially not the umbrella pistol.” “It only fires confetti.” “It fired a fork last Tuesday.” “The fork was already inside.” “Bill.” “Fine. No weapons. Can I bring wrapping paper?” “Why?” “If we’re caught, we say we’re delivering presents.” “That is almost intelligent.” “Thank you.” “Except we’ll be carrying them out.” Bill paused. “We say they’re returns.” “All forty-seven?” “Children are notoriously dissatisfied.” Molly headed for the door. Bill hurried after her. “One final question.” “What?” “Can we keep the cake?” “No.” “Half the cake?” “No.” “A slice?” “Only if we escape unseen.” Bill grinned. “At last, a motive.”
created_at
"2026-07-09T17:56:06.411Z"
7/9/2026, 5:56:06 PM
model
"openai/gpt-5.6-sol-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
796
moderation_latency
(null)
generation_time
21466
tokens_prompt
77
tokens_completion
1560
native_tokens_prompt
66
native_tokens_completion
1028
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
0.03117
router
(null)
provider_responses
0
endpoint_id
"a54c5de0-89bf-4ad7-a212-cf977eed918a"
id
"resp_094d84fca55149fa016a4fe0b67a3c8195bd6d1fa9a12de48e"
is_byok
false
latency
199
model_permaslug
"openai/gpt-5.6-sol-20260709"
provider_name
"OpenAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1783619766-letnYzBpVwdSz8WhhH4L"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783619766-jf8wvEe45tojsEMegqWV"
upstream_id
"resp_094d84fca55149fa016a4fe0b67a3c8195bd6d1fa9a12de48e"
total_cost
0.03117
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
678
0.00%
Dialogue to Total Word Ratio
Ratio: 84.19%, Deviation: 14.19%
neededClean
false
wordsTotal
683
wordsDialogue
575
0.0021%