NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37HNBR7DPX7ZH95P3PFQ16P
Run Details
Model:
GPT-6 Sol
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
17.5s
Cost:
$0.007182
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
9.61%
Matches word count
(details)
28.93%
Dialogue to Total Word Ratio
(details)
19.2669%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% 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 drawing across the café table. A purple rectangle sat beneath a furious yellow sun. “That’s the house?” Bill asked. “That’s the bouncy castle. The house is the brown smudge behind it.” “Why is there a dragon?” “Birthday girl’s artistic interpretation of her father.” Bill studied the drawing. “He looks alert.” “He’s barbecuing for thirty children. By noon, he’ll be begging for a dragon to eat him.” Molly tapped a cluster of red squares. “Presents. They’re piling them beside the patio door.” “All of them?” “All the ones guests bring.” “That seems greedy.” “We’re thieves, Bill. Greedy is our dental plan.” Their waiter delivered two coffees and an invoice Molly had requested for “surveillance stationery.” Bill waited until he left. “How do we get past the parents?” “Disguise. I’m the magician. You’re my assistant.” “I thought we agreed I wouldn’t be an assistant again.” “You said you wouldn’t be a magician’s rabbit again. This is a promotion.” Bill looked at the rabbit ears poking from Molly’s tote. “Why are those here?” “Contingency.” Outside, a woman shepherded three children past the window. One child wore a cape and carried a foil sword. Bill watched him with professional unease. At parties, children noticed everything adults overlooked, particularly strangers holding sacks. “What’s the exit plan?” he asked. “We load the presents into my car while everyone watches the magic show.” “You can do magic?” “I can make thirty parents look at their phones simultaneously.” “That isn’t magic.” “It will be when I tell them their children are volunteering.” Bill rubbed his face. “What if someone asks where we’re taking the gifts?” “We say we’re moving them indoors before the rain.” “The forecast says sunshine.” Molly flipped the drawing over. On the back, a blue cloud bore the words MAYBE RAIN in careful block capitals. “Whose writing is that?” “Mine. I’m establishing reasonable doubt.” A party horn sounded from a nearby car. Both thieves flinched. Molly folded the paper and tucked it away as if it contained state secrets rather than a dragon with impressive eyebrows. Bill lowered his voice. “And afterward? We open everything?” “No. We deliver it.” “To whom?” “The children’s hospital. Remember the charity auction? This is the collection party.” Bill stared at her. “You said we were stealing all the presents.” “We are. From the party. Temporarily. The organiser hired us to stage a dramatic pickup.” “Why didn’t you tell me?” “You’d have practiced a speech.” “I have practiced a speech.” “Exactly.” Bill drew himself up. “Did you at least tell the organiser about the rabbit costume?” “No,” Molly said. “That surprise is for the children.” By eleven, Bill stood outside the garden gate in a waistcoat, rabbit ears, and an expression suggesting he had lost a bet with the universe. Molly rang the bell. A small girl opened it, examined Bill, and held out her hand. “Ticket,” she said. Molly blinked. “We’re the entertainment.” “Then you definitely need one.” Bill produced a chocolate coin from behind the girl’s ear. She inspected it, pocketed it, and waved them through. “Excellent,” Molly whispered. “Twenty more to bribe,” Bill whispered back.
created_at
"2026-09-23T16:30:27.598Z"
9/23/2026, 4:30:27 PM
model
"openai/gpt-6-sol-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3858
moderation_latency
(null)
generation_time
17509
tokens_prompt
75
tokens_completion
926
native_tokens_prompt
66
native_tokens_completion
705
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
0.007182
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790181027-GoRsDMZiSSdDYYHI7sct"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790181027-nvWd1s2q9EWaBZ2RehXF"
upstream_id
"resp_08c80e0d6c767001016ab3fea3d2d087d1a7edebfc4a937939"
provider_responses
0
endpoint_id
"1cfc7d9d-4404-4b8e-9ee4-58ae45c9dcd4"
id
"resp_08c80e0d6c767001016ab3fea3d2d087d1a7edebfc4a937939"
is_byok
false
latency
1704
model_permaslug
"openai/gpt-6-sol-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.007182
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
9.61%
Matches word count
n/a
neededClean
false
words
522
28.93%
Dialogue to Total Word Ratio
Ratio: 54.98%, Deviation: 4.98%
neededClean
false
wordsTotal
522
wordsDialogue
287
19.2669%