NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37HTVZ7NQQ6GDF9X2D80XAN
Run Details
Model:
GPT-6 Sol
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
15.0s
Cost:
$0.006952
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
99.74%
Matches word count
(details)
48.30%
Dialogue to Total Word Ratio
(details)
74.0204%
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. It showed a house, a swimming pool, and twelve stick figures with enormous triangular hats. “Are those security guards?” Bill asked. “Children. The guards are wearing party hats too, but they have necks.” Bill leaned closer. “Why does that one have six arms?” “That’s the birthday girl. She drew the map herself.” “You got intelligence from a seven-year-old?” “She’s eight on Saturday. Show some respect.” Outside, rain rattled against the café window. Molly weighed down the drawing with a sugar bowl while Bill studied the enormous rectangle labeled PRESENTS!!! It sat beside a smaller rectangle labeled BROCCOLI DANGER. “Where do they keep the presents?” he asked. “Under the banner. Dining room.” “And the broccoli?” “I didn’t ask. She seemed upset.” Molly had spent three weeks cultivating a connection with the birthday girl, mostly by losing repeatedly at miniature golf. She knew the layout, the guest list, and the exact moment the magician would produce a rabbit. Unfortunately, the invitation insisted all adults wear costumes. “I’ll be a pirate,” Bill said. “No pirates. Her uncle’s a pirate, and he knows every other pirate in town.” “Clown?” “Her aunt’s a clown.” “Wizard?” “The magician will take it personally.” Bill sighed. “What’s left?” Molly slid him a photograph of a smiling cartoon vegetable. He stared. “That’s a courgette.” “You’ll be unforgettable.” “That’s bad for a thief.” “You’ll also be unrecognizable.” The café owner arrived with their toast, took one look at the drawing, and asked whether they were organizing a party. Molly said yes. Bill said no. The owner retreated. Molly lowered her voice. “We arrive during the magic show, collect the presents, and leave through the side gate.” “Carry them how?” “In the vegetable costume. It’s roomy.” Bill examined the photograph again. “Where do my arms go?” “Inside.” “How do I carry anything?” Molly paused. Rain drummed harder. She turned the photograph over, as if the answer might be printed on the back. “You’re not the bag,” she admitted. “You’re the distraction.” “What kind of distraction?” “A dancing courgette.” “I don’t dance.” “You’ve danced at weddings.” “I’ve fallen over at weddings.” “Exactly.” Bill pushed his toast away. “I want half.” “You already get half.” “I want the half without glitter on it.” Across the table, Molly unfolded a second document: a spreadsheet estimating the resale value of toy dinosaurs, craft kits, and one electric scooter. Bill read it with growing discomfort. “Those are birthday presents,” he said. “That is the broad premise.” “For a little girl who trusted you with her map.” Molly looked at the drawing. In the corner, beneath a purple sun, the birthday girl had written MOLLY IS MY BEST FRIEND in wobbly capitals. “Oh,” Molly said. Bill nodded. They sat quietly until Molly folded the spreadsheet and tucked it away. “We could steal the magician’s rabbit,” Bill suggested. “No.” “His wallet?” “No.” “His secret?” Molly smiled. “That,” she said, “might be worth a party invitation.”
created_at
"2026-09-23T16:33:28.049Z"
9/23/2026, 4:33:28 PM
model
"openai/gpt-6-sol-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3011
moderation_latency
(null)
generation_time
14941
tokens_prompt
75
tokens_completion
897
native_tokens_prompt
66
native_tokens_completion
682
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
0.006952
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790181208-hc8Ehn6QEWxjF4nK6nCg"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790181208-0UePi3fbxAqda0rSeGMs"
upstream_id
"resp_0326f8e927e013c3016ab3ff582e5087d19d765ff9eb1244c8"
provider_responses
0
endpoint_id
"1cfc7d9d-4404-4b8e-9ee4-58ae45c9dcd4"
id
"resp_0326f8e927e013c3016ab3ff582e5087d19d765ff9eb1244c8"
is_byok
false
latency
1387
model_permaslug
"openai/gpt-6-sol-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.006952
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
99.74%
Matches word count
n/a
neededClean
false
words
496
48.30%
Dialogue to Total Word Ratio
Ratio: 46.18%, Deviation: 3.82%
neededClean
false
wordsTotal
498
wordsDialogue
230
74.0204%