NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37HVS8192TMW7NEMREZV7JA
Run Details
Model:
GPT-6 Sol
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
19.5s
Cost:
$0.007762
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.03%
Matches word count
(details)
39.73%
Dialogue to Total Word Ratio
(details)
19.8821%
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. A pink smear marked the house. A blue square marked the bouncy castle. Someone had drawn a dinosaur in the driveway. “Is that security?” Bill asked. “That’s a dinosaur.” “I can see that. Is it working security?” “It’s a drawing, Bill.” “Good. I’m excellent with drawings.” Molly tapped the pink smear. “Saturday. Four o’clock. Twelve children, approximately fourteen adults, and one enormous pile of presents beside the cake table.” “How enormous?” “Properly enormous. I had to take a second lap past the window to see around it.” “And you’re sure they’re presents?” “They’re wrapped, tied with bows, and labelled ‘Happy Birthday, Poppy.’” “Could be bricks.” “For a seven-year-old?” “Some parents are ambitious.” A waiter set down two coffees. Molly waited until he had gone. “We go in during the magician’s act,” she said. “Everyone watches the disappearing rabbit. We make the presents disappear too.” “That sounds like competition. Magicians hate competition.” “The magician won’t notice.” “What if the rabbit does?” Molly folded her arms. “Do you want to help or not?” “I do. I just want to know whether we’re robbing a child with a rabbit as a witness.” “No witnesses. We’ll wear costumes.” Bill brightened. “I have a pirate hat.” “We’re going as parents.” “Parents can be pirates.” “Not these parents. I watched three of them argue over whether glitter is biodegradable.” Bill sipped his coffee. “What do I wear, then?” “Normal clothes. Carry a casserole.” “I don’t own a casserole.” “You don’t wear it. You bring one.” “What’s in it?” “Nothing.” “An empty casserole makes us look suspicious.” “Fine. Put potatoes in it.” “Cooked?” Molly closed her eyes. “Surprise me.” She sketched an arrow from the front door to the cake table. Bill leaned over the map, leaving a ring of coffee across the bouncy castle. “We collect the presents in these laundry bags,” she said. “Then we leave through the kitchen.” “Why the kitchen?” “The back door opens onto the alley.” “What if someone asks who we are?” “Poppy’s cousins.” “Both of us?” “Yes.” “Do we look related?” “We look equally tired.” “That’s fair.” Bill studied the map. “Where’s the bathroom?” “Why?” “Because if we’re pretending to be parents, one of us should ask where it is immediately.” Molly nodded despite herself. “Good thinking.” “And we should know Poppy’s age.” “Seven.” “Favourite animal?” “I don’t know.” “School?” “I don’t know.” “Best friend?” “Bill, we’re stealing presents, not attending a parent-teacher conference.” “I want a convincing character.” “Then be a cousin who doesn’t remember things.” “That’s my strongest role.” He pointed at the cake table. “One problem.” “What?” “Poppy might see us carrying her presents away.” Molly stared at the map. Outside, a child dragged a balloon along the pavement, laughing whenever it bounced. “We could wait until after she opens them,” Bill said. “Then they won’t be wrapped.” “We’re stealing toys, not paper.” Molly considered this. “After cake, then. During cleanup.” “And we take only the presents nobody claims?” “That’s not how stealing works.” “It might be how we start.” Molly rolled up the map. “Bring the potatoes.” Bill grinned. “Cooked?” “Please, Bill. Let me have one surprise.”
created_at
"2026-09-23T16:33:58.033Z"
9/23/2026, 4:33:58 PM
model
"openai/gpt-6-sol-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3207
moderation_latency
(null)
generation_time
19502
tokens_prompt
75
tokens_completion
1004
native_tokens_prompt
66
native_tokens_completion
763
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
0.007762
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790181238-AxoCVvfsVAX0OnpyDJLY"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790181238-B4FRF3PK6jP7VfRK4d8w"
upstream_id
"resp_05fd99179c2fa7e2016ab3ff76271c87d1808ede33bcbb250f"
provider_responses
0
endpoint_id
"1cfc7d9d-4404-4b8e-9ee4-58ae45c9dcd4"
id
"resp_05fd99179c2fa7e2016ab3ff76271c87d1808ede33bcbb250f"
is_byok
false
latency
1127
model_permaslug
"openai/gpt-6-sol-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.007762
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
0.03%
Matches word count
n/a
neededClean
false
words
530
39.73%
Dialogue to Total Word Ratio
Ratio: 74.30%, Deviation: 4.30%
neededClean
false
wordsTotal
533
wordsDialogue
396
19.8821%