NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX40RVZVQJ7J9FWJJ8CRWK56
Run Details
Model:
GPT-5.6 Sol (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
53.6s
Cost:
$0.140550
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.56%
Dialogue to Total Word Ratio
(details)
99.7794%
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
Rain polished the bakery window as Molly spread a crayon map across the counter. Bill squinted at its balloons, dragon cake, and ominously cheerful clown. “At three o’clock, the children begin musical statues. At three-oh-five, the parents argue about whether Jasper moved. That’s our window.” Molly tapped the gift table with a blue crayon. “Our window is five minutes wide and guarded by a clown. Molly, I was bitten by one. Emotionally, I never recovered.” Bill shuddered so violently that a tray of buns rattled. “He didn’t bite you. He made balloon trousers for you, and you sat on them. The bang startled him into dropping his teeth.” “False teeth are still teeth. Also, those trousers were structurally unsound and surprisingly judgmental around the hips.” Molly circled the back door, then drew a moustache on the clown. “We enter as entertainers, wheel every present into our van, and leave before the candles are lit. Simple, elegant, educational.” “Entertainers doing what? My juggling resembles an argument with gravity, and your magic act once produced a pigeon that refused to acknowledge you.” Molly folded her arms. Thunder applauded. “We’ll be gift inspectors. Parents adore officials, especially officials carrying clipboards and looking disappointed by ribbon quality.” “Children don’t adore officials. Children smell uncertainty. Last Christmas, a six-year-old interrogated me until I confessed to borrowing your hairbrush.” Molly stared at his shiny bald head. Bill carefully adjusted his hat. “That confession raises several questions, none relevant to crime.” “What if we crawl through the bouncy castle tunnel? We emerge beside the presents, unseen, like ninjas with static electricity.” She considered this, chewing the crayon until blue flecked her smile. “The tunnel ends in the ball pit. We’d surface covered in plastic balls, juice, and whatever children mysteriously secrete.” “Confidence. Children secrete confidence. And jam, apparently, even when no jam has been served.” A timer pinged. Both thieves jumped; cupcakes emerged. “Backup plan: we pose as responsible adults, walk in through the front, compliment the decorations, and take only presents labeled ambiguously.” “Ambiguously? They all say happy birthday, Sophie. Unless one says happy birthday, strangers, please steal this.” Molly lifted a cupcake and inspected its pink icing like forensic evidence. “Fine. We take the untraceable ones: art supplies, socks, books. We leave anything personalized, noisy, or morally complicated.” “Morally complicated? We’re robbing children. The morality has already removed its shoes, telephoned a cab, and gone home.” Rain filled the silence. Molly rolled up the map. “You’re right. New plan: we break in, add two presents to every pile, and steal the clown’s false teeth.” “At last, a heist with standards. What presents?” Molly opened the cupboard, revealing forty knitted octopuses in tiny hats. “My aunt’s craft phase. They’re unsellable, indestructible, and technically machine washable.” “And the children will love them?” Bill picked up an octopus. Its stitched smile seemed to know his tax history. “They’ll learn to. Load the van.” “After we rescue those teeth.” Molly grinned. Thunder applauded, and somewhere across town, a clown sneezed.
created_at
"2026-07-09T18:03:22.502Z"
7/9/2026, 6:03:22 PM
model
"openai/gpt-5.6-sol-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
4465
moderation_latency
(null)
generation_time
52823
tokens_prompt
77
tokens_completion
1412
native_tokens_prompt
66
native_tokens_completion
4674
native_tokens_completion_images
0
native_tokens_reasoning
3970
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
0.14055
router
(null)
provider_responses
0
endpoint_id
"a54c5de0-89bf-4ad7-a212-cf977eed918a"
id
"resp_04e74b4a4b6b7be0016a4fe26a9cec8195ae2733544d480a41"
is_byok
false
latency
260
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-1783620202-QbDr8QIMIYFMKJpvc7pG"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783620202-B1hYH7y2jnZ2icc1Lqfm"
upstream_id
"resp_04e74b4a4b6b7be0016a4fe26a9cec8195ae2733544d480a41"
total_cost
0.14055
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
100.00%
Matches word count
n/a
neededClean
false
words
501
99.56%
Dialogue to Total Word Ratio
Ratio: 70.30%, Deviation: 0.30%
neededClean
false
wordsTotal
505
wordsDialogue
355
99.7794%