NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37HQP0YD34QTMTTYGYD859H
Run Details
Model:
GPT-6 Sol
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
12.4s
Cost:
$0.006642
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
99.38%
Matches word count
(details)
5.67%
Dialogue to Total Word Ratio
(details)
52.5245%
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 floor plan across the café table, trapping its corners beneath two sugar shakers and a muffin. At the next table, a toddler banged a spoon against a cup. Molly waited for the noise to stop, then tapped a red circle in the plan. “Four o’clock. Cake comes out here.” Bill leaned closer. “You drew that circle around a bouncy castle.” “The cake is beside it.” “Then why’s there a crown on the castle?” “That’s the princess.” “Is she a threat?” “She’s six, Bill.” “She has a crown.” Molly took back her pen. The party was at the community hall on Saturday. Twenty children, one magician, and a present table positioned just inside the side door. She had learned this from the invitation stuck to her neighbour’s fridge. She had also learned that her neighbour expected her to bring potato salad. Bill studied the plan. “Are we going as parents?” “No. You look frightened of children.” “I am frightened of children.” “You’ll wear the rabbit costume.” He pushed the plan away. “The magician has a rabbit.” “The magician has a stuffed rabbit. You’re a surprise guest.” “Surprise to whom?” “To the magician, for a start.” Molly had found the costume in a charity shop window that morning. Its ears stood straight, its paws were enormous, and the zip had a tendency to catch. She had spent forty minutes inside it before she could pull it off again. Bill watched her place the receipt on the table and turn it over, revealing a sketch of the side entrance. “Once you’re in, you distract them,” she said. “I take the presents out to the van.” “All the presents?” “That’s the job.” “What if one’s a bicycle?” “We’ll make two trips.” “What if there’s a child on it?” “Wait until they get off.” The toddler at the next table dropped the spoon. Bill picked it up, wiped it on a napkin, and handed it back. The toddler stared at him, then knocked the cup over. Juice spread across Molly’s plan, turning the side entrance into a blue lagoon. Molly lifted the paper by two corners. “We can still read it.” Bill pointed to the invitation, which had floated loose. “Why does it say ‘No presents, please’?” Molly blinked. She turned the damp card around. Beneath a glittery drawing of a dinosaur, in purple capitals, were the words NO PRESENTS, PLEASE. DONATIONS TO THE ANIMAL SHELTER WELCOME. For a moment, neither thief spoke. The toddler resumed banging the spoon. “Well,” said Bill, “the rabbit costume’s appropriate.” “We’re not stealing donations from an animal shelter.” “I meant we could give them the costume.” Molly imagined telling her neighbour why Bill owned it. She folded the soggy floor plan, tucked it into her bag, and finished her coffee. “Saturday,” she said, “you’re coming with me to buy potato salad.” Bill looked relieved. “Can I wear my own clothes?” “Yes.” “Good. The rabbit paws won’t fit through a shop door.”
created_at
"2026-09-23T16:31:43.654Z"
9/23/2026, 4:31:43 PM
model
"openai/gpt-6-sol-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2477
moderation_latency
(null)
generation_time
12370
tokens_prompt
75
tokens_completion
804
native_tokens_prompt
66
native_tokens_completion
651
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
0.006642
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790181103-KzKUoYHvLL2MKT3IcFy7"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790181103-k3YwoDi2qXPboG1ffiUP"
upstream_id
"resp_022f0bb83259dc91016ab3feefc4a487d186d5ac44bad72a08"
provider_responses
0
endpoint_id
"1cfc7d9d-4404-4b8e-9ee4-58ae45c9dcd4"
id
"resp_022f0bb83259dc91016ab3feefc4a487d186d5ac44bad72a08"
is_byok
false
latency
754
model_permaslug
"openai/gpt-6-sol-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.006642
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
99.38%
Matches word count
n/a
neededClean
false
words
495
5.67%
Dialogue to Total Word Ratio
Ratio: 37.58%, Deviation: 7.58%
neededClean
false
wordsTotal
495
wordsDialogue
186
52.5245%