NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX41N00TY6XDBYKTE2NM95A0
Run Details
Model:
GPT-5.6 Sol (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
56.1s
Cost:
$0.149790
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 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 map across the dashboard while rain polished the windshield and the hired clown vanished through the community hall’s glittering doors ahead. “The birthday boy’s name is Alfie, the cake resembles a dinosaur, and every present waits beside the bouncy castle under a balloon arch indoors today.” Bill studied the map upside down, nodding with the grave authority of a surgeon who had accidentally enrolled in balloon-animal school that morning before breakfast. “We enter disguised as entertainers, locate the gift table, swap the parcels for empty boxes, and escape during the magician’s dramatic smoke bomb grand finale.” From a shopping bag, Molly produced two costumes: a fairy princess with steel-toed boots and a pirate outfit apparently designed for a nervous indoor dachshund. “I ordered adult medium. This label says ages six to eight, and the trousers include a tail hole shaped like a heart for ventilation purposes.” Molly held the pirate tunic against him. It reached his ribs. The plastic hook, however, was enormous enough to tow a modest fishing boat ashore. “Perfect. Nobody questions a pirate with poor tailoring. They assume authenticity, vitamin deficiencies, or a complicated backstory involving parrots and unfavorable exchange rates abroad lately.” He squeezed into the costume, splitting one seam with a sound like a tiny pistol. Molly solemnly pinned the rupture using a unicorn badge very carefully. “Now the difficult part begins. Children can smell dishonesty. Last Christmas, my niece identified counterfeit snow before I had even plugged the machine in upstairs.” Molly opened her briefcase. Instead of tools, it contained juice boxes, wet wipes, three yo-yos, and a laminated certificate declaring her Official Assistant Fairy Today. “Children respect paperwork. If challenged, I distract them with stickers. You roll the gifts out through the kitchen in that laundry cart by the bins.” Bill peered toward the hall. A small girl in a tiara stood at the window, watching them with the pitiless concentration of airport security staff. “We have been spotted. Smile naturally, Molly. Not your bank-robbery smile; the warm one suggesting kindness, puppets, and absolutely no outstanding warrants anywhere on Earth.” Molly waved. The girl tapped the glass, then pointed sternly at Bill’s hook. Bill waved it back, accidentally snagging and beheading the car’s fuzzy dice. “She’s onto us. That child has seen things. Probably bedtime after nine. Possibly unsupervised access to detective cartoons and fizzy orange on Saturday mornings alone.” The hall doors burst open. The tiara girl marched across the puddles, carrying a clipboard and wearing an expression too managerial for someone missing incisors. “Stay calm. According to her clipboard, the clown quit, the magician is crying, and our pirate must organize musical chairs immediately without lawsuits or fatalities.” Bill’s face brightened with professional pride. Plans changed; chaos remained. He climbed from the car, his tiny trousers creaking, and saluted the girl with grandeur. “New plan: we run the party brilliantly, earn everyone’s trust, load the presents afterward, then discover we’re too exhausted and morally improved to steal anything.”
created_at
"2026-07-09T18:18:44.131Z"
7/9/2026, 6:18:44 PM
model
"openai/gpt-5.6-sol-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
4180
moderation_latency
(null)
generation_time
55871
tokens_prompt
77
tokens_completion
1513
native_tokens_prompt
66
native_tokens_completion
4982
native_tokens_completion_images
0
native_tokens_reasoning
4337
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
0.14979
router
(null)
provider_responses
0
endpoint_id
"a54c5de0-89bf-4ad7-a212-cf977eed918a"
id
"resp_0c532fddf4c8efe4016a4fe604365481909df09b6e7479eb23"
is_byok
false
latency
252
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-1783621124-LZhQ63yf56v5YdwcAVrr"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783621124-z4X0686tJzJomXvo5HkS"
upstream_id
"resp_0c532fddf4c8efe4016a4fe604365481909df09b6e7479eb23"
total_cost
0.14979
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: 49.70%, Deviation: 0.30%
neededClean
false
wordsTotal
505
wordsDialogue
251
99.7794%