NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX4112DTV5EH5H240647THZP
Run Details
Model:
GPT-5.6 Sol
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
17.8s
Cost:
$0.022890
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.96%
Dialogue to Total Word Ratio
(details)
0.4815%
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 spread the birthday invitation across the café table, pinning its corners with a sugar bowl, a chipped mug, and Bill’s elbow. “Move your arm. You’re crushing the unicorn.” “I distrust unicorns,” Bill said. “Horses shouldn’t carry knives.” “It’s a horn.” “That’s what they want you to think.” Around them, spoons clinked and milk steamed. Nobody noticed Molly draw a floor plan in spilled sugar. The Happy Acorn Play Barn occupied a converted bank, which Molly considered poetic. Once, adults had guarded money there. Now, children guarded presents with sticky fingers and unpredictable screaming. “The party starts at two,” Molly said. “Thirty guests, aged five to seven.” “Dangerous age.” “You say that about everyone.” “Everyone is dangerous until proven asleep.” Bill studied the plan. The present table stood inside the old vault, beneath a papier-mâché dragon. Entry required passing reception, the parents’ lounge, and an entertainer named Professor Pickle. “What’s Professor Pickle’s specialty?” he asked. “Balloon animals and close-up magic.” “A rival professional.” “You once made a pigeon disappear.” “It escaped.” “With your wallet.” Molly outlined the scheme. They would arrive disguised as children’s entertainers. Bill would distract the guests with juggling while Molly entered the vault, loaded the presents into collapsible laundry sacks, and exited through the emergency door. Bill’s face developed the solemnity it reserved for catastrophe and mayonnaise. “I can’t juggle.” “You told me you could.” “I can juggle facts.” “No, you confuse facts. That’s different.” A waitress approached, and Molly transformed the sugar blueprint into an innocent dune. The waitress deposited two muffins. Bill watched her leave, then whispered as though informing national intelligence. “We need a better disguise.” “We have wigs.” “My wig makes me look like a haunted mop.” “Perfect. Children love haunted things.” “They love dinosaurs. I could be a dinosaur.” “You can’t reach a vault keypad with tiny arms.” Bill conceded this by attacking his muffin. Outside, rain polished the pavement. A delivery van passed bearing the Happy Acorn logo: a squirrel grinning with too many teeth. Molly photographed it, already calculating uniforms, schedules, and whether Bill could convincingly impersonate anyone employed. “The presents leave at four,” she said. “We’ll be gone by three.” “What if a child recognizes us?” “From where?” Bill lowered his voice. “The aquarium incident.” “Those children thought you were a seal.” “I was wearing a dinner jacket.” “You were also in the seal enclosure.” He sighed. Molly gathered the invitation and brushed the sugar away. Her finger paused over the guest of honor’s name: Poppy Wainwright. Beside it, in glitter pen, someone had added NO CLOWNS. “Well,” Molly said, “that simplifies costumes.” Bill brightened. “Dinosaurs?” “No.” “Magicians?” “No.” “Tax inspectors?” Molly stared at him. “Children fear them instinctively,” he explained. Before she could answer, the café door opened. A woman entered carrying three towers of wrapped presents, each topped with a silver bow. Molly and Bill turned together, their professional attention snapping into place. The woman set them beside the counter and asked for directions to the Happy Acorn. Bill smiled. “Change of plan?” Molly watched a bow shimmer. “Absolutely. We help her carry them.” “And then?” “We discover whether honesty creates opportunities.” Bill rose, collected two towers, and staggered beneath them. “If it doesn’t?” Molly lifted the third. “Then we’re already holding the presents.”
created_at
"2026-07-09T18:07:51.235Z"
7/9/2026, 6:07:51 PM
model
"openai/gpt-5.6-sol-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
651
moderation_latency
(null)
generation_time
17570
tokens_prompt
77
tokens_completion
1088
native_tokens_prompt
66
native_tokens_completion
752
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
0.02289
router
(null)
provider_responses
0
endpoint_id
"a54c5de0-89bf-4ad7-a212-cf977eed918a"
id
"resp_0810a625b7f27ec2016a4fe37755e481979e8218a62593e2bf"
is_byok
false
latency
202
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-1783620471-ZYaoCihKrxlWj6emTigr"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783620471-pQbyxPZ1hP0u1mqu0jYA"
upstream_id
"resp_0810a625b7f27ec2016a4fe37755e481979e8218a62593e2bf"
total_cost
0.02289
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
548
0.96%
Dialogue to Total Word Ratio
Ratio: 39.64%, Deviation: 9.64%
neededClean
false
wordsTotal
550
wordsDialogue
218
0.4815%