NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX418R9A52VR4BMYCFXJHF7F
Run Details
Model:
GPT-5.6 Terra
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
12.9s
Cost:
$0.013380
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0012%
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 crouched behind the inflatable castle, peering through a curtain of plastic palms at the garden beyond. Balloons bobbed over the fence like surveillance drones. Somewhere near the patio, a magician was making a rabbit disappear to sustained shrieking. Bill adjusted his black ski mask, which had a glittery unicorn horn stitched to the forehead. “Why am I wearing this?” he whispered. “Because it was in the disguise bag.” “It was in your niece’s dressing-up box.” “Operationally identical.” Bill studied the target: a picnic table sagging under wrapped parcels, gift bags, and one enormous cardboard box decorated with dinosaurs. “I still think we should wait until cake.” “We wait until cake, the children become sticky and unpredictable.” “They’re already unpredictable. One of them just tried to eat a balloon.” “Exactly. We move during the magician’s finale. All eyes on the rabbit. You take the presents, I create a diversion.” Bill looked at her. “What kind of diversion?” Molly produced a party whistle, three smoke pellets, and a laminated sheet of temporary tattoos. “Chaos with a festive theme.” “You brought smoke pellets to a six-year-old’s birthday?” “They’re pink.” “That does not make them safer.” “It makes them coordinated.” A child in a cape ran past the hedge, yelling, “I’m a laser shark!” and fired a foam dart into Bill’s ear. Bill winced. “We’re robbing tiny warlords.” “Focus. The gift table is ten paces from the back gate. We bag everything, vault the fence, and meet the van by the recycling bins.” “The van says ‘Molly’s Mobile Pet Grooming.’” “Perfect cover.” “It has a painted poodle wearing sunglasses.” “People trust a poodle in sunglasses.” The magician announced, “For my next trick, I need absolute silence!” Forty children screamed at once. Molly nodded sharply. “Now.” They slipped from behind the castle. Bill pushed a wheeled laundry hamper toward the present table while Molly marched into the crowd with the confidence of a woman delivering emergency cupcakes. “Attention, junior citizens!” she cried. “Who wants free tattoos?” The children turned as one. Bill froze. “That’s your diversion?” “Sell it!” Molly hissed. Molly slapped a temporary pirate skull onto the nearest child’s forehead. “Congratulations, you’re now captain of the snack table!” A tidal wave of capes, crowns, and sugar-fueled limbs surged toward her. Bill reached the presents. He began loading them into the hamper with frantic efficiency. “Pink unicorn,” he muttered. “Board game. Suspiciously heavy robot. Gift bag shaped like a llama. Why do children get so much stuff?” “Because adults are weak!” Molly called, pinned against the patio wall by three girls demanding glitter dolphins. Bill lifted the dinosaur box. It rattled. “Please don’t be a dinosaur,” he whispered. The box burst open. A small boy popped up wearing night-vision goggles and a plastic sheriff’s badge. “Caught you!” he shouted. Bill stared. “Are you the birthday boy?” “I’m Toby. This is my security bunker.” “Very impressive bunker.” “Those are my presents.” Bill considered several responses, rejected all of them, and offered the boy a foam dart. “Would you like to join our criminal enterprise?” “Toby!” Molly barked, suddenly beside him. “Your magician is turning your dad into a frog!” Toby gasped and sprinted away. Bill shoved the dinosaur box into the hamper. “That was low.” “It was technically plausible.” They raced for the gate. Behind them, the magician’s rabbit leapt from his hat, stole a sausage roll, and vanished beneath the table. Bill hauled the hamper over the fence. Molly climbed after him, her party whistle clenched between her teeth. As they landed in the alley, Bill opened a gift bag. Inside was a glittery picture frame, a packet of socks, and a handwritten card. He read it aloud. “‘Dear Toby, hope you have the best birthday ever. Love, Auntie Karen.’” Molly sighed. “Fine. We return the sentimental ones.” “And the socks?” “Keep the socks. We have standards.”
created_at
"2026-07-09T18:12:02.993Z"
7/9/2026, 6:12:02 PM
model
"openai/gpt-5.6-terra-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
510
moderation_latency
(null)
generation_time
12687
tokens_prompt
77
tokens_completion
1179
native_tokens_prompt
66
native_tokens_completion
881
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
0.01338
router
(null)
provider_responses
0
endpoint_id
"3cc89000-ae34-4dde-9c7d-5ce009c8af0b"
id
"resp_0e1182d504e1dd08016a4fe47318608190875545fda8e31bbd"
is_byok
false
latency
191
model_permaslug
"openai/gpt-5.6-terra-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-1783620723-BDgftIAsNRlfdAQxAkmP"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783620723-lG8IbpFR0XXiwnMIt0BC"
upstream_id
"resp_0e1182d504e1dd08016a4fe47318608190875545fda8e31bbd"
total_cost
0.01338
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
0.00%
Matches word count
n/a
neededClean
false
words
643
0.00%
Dialogue to Total Word Ratio
Ratio: 44.60%, Deviation: 14.60%
neededClean
false
wordsTotal
648
wordsDialogue
289
0.0012%