NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX41NB34K31DQHR0J1ZW61HC
Run Details
Model:
GPT-5.6 Sol (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
20.9s
Cost:
$0.031320
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
46.86%
Dialogue to Total Word Ratio
(details)
23.4281%
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
Molly spread the crayon-smeared floor plan across the café table, pinning one corner with a sugar bowl and another with Bill’s custard tart. “Right,” she said. “At three o’clock, the magician arrives.” “Professional rival?” “Distraction.” “I’ve always distrusted magicians. Nobody needs that many scarves unless they’re concealing cutlery.” “At three fifteen, the children gather in the garden. At three twenty, we enter through the kitchen dressed as entertainers.” “What sort?” “Balloon sculptors.” Bill lowered his fork. “Molly, I can’t sculpt balloons.” “You made that dachshund.” “That was meant to be a sword.” “It had ears.” “They were defensive handles.” Molly tapped the plan. “We cross the kitchen, pass the pantry, enter the hall, and reach the present room.” “Present room,” Bill whispered reverently. “Like a bank vault, only with dinosaurs and glitter.” “Exactly. We fill these sacks, leave through the laundry door, and vanish before anyone notices.” Bill examined the sacks beneath his chair. Each was printed with HAPPY CHRISTMAS FROM DORSET POTATOES. “Subtle.” “They were discounted.” “It’s June.” “That’s why.” A waiter passed. Molly smiled until he was gone. Bill leaned closer. “What about security?” “Two grandmothers.” “Armed?” “Handbags.” His face paled. “So, yes.” “One watches the children. The other guards the buffet.” “What’s on the buffet?” “Focus.” “I need to understand the terrain.” “Sausage rolls, sandwiches, jelly, cake.” “What kind of cake?” “Chocolate.” “With ganache?” “Bill.” “This affects my route.” “You are not stopping for cake.” “Then why become a criminal?” Molly drew a dotted line in purple crayon. “Here’s the problem. The family dog sleeps outside the present room.” “Breed?” “Pomeranian.” Bill laughed, then saw her expression. “Oh. Veteran?” “Bit a postman, a plumber, and Father Christmas.” “Maybe Father Christmas had it coming.” “We neutralize it with chicken.” “Cooked?” “Obviously.” “Seasoned?” “It’s a guard dog, not a restaurant critic.” Bill considered this. “Could disguise myself as a larger Pomeranian.” “No.” “I’ve got the coat.” “No.” “The tail is convincing.” “The tail frightened a traffic warden.” “Proof of quality.” Molly folded the plan. “Once inside, take wrapped presents only. No bicycles, no dollhouse, no suspiciously ticking packages.” “What if a ticking package is an educational clock?” “Leave it.” “What if we accidentally steal disappointing presents?” “All presents are disappointing. That’s adulthood.” Bill finished his tart and frowned at the invitation. “Molly, it says guests should bring donations for the children’s hospital.” She snatched it, read twice, and sighed. “So the presents aren’t for one spoiled child?” “No.” “They’re for sick children?” “Yes.” A long silence settled between them. Bill pushed back his chair. “We can’t steal those.” “Of course not.” “Good.” “We’ll steal the magician’s van.” Bill brightened. “Will there be scarves?” “Thousands.” “And possibly cake?” “Get the sacks.”
created_at
"2026-07-09T18:18:55.486Z"
7/9/2026, 6:18:55 PM
model
"openai/gpt-5.6-sol-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
5151
moderation_latency
(null)
generation_time
20724
tokens_prompt
77
tokens_completion
1132
native_tokens_prompt
66
native_tokens_completion
1033
native_tokens_completion_images
0
native_tokens_reasoning
341
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
0.03132
router
(null)
provider_responses
0
endpoint_id
"a54c5de0-89bf-4ad7-a212-cf977eed918a"
id
"resp_04c91eba4aa8be6c016a4fe610d4188195b70c5a60a36bb1b1"
is_byok
false
latency
1528
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-1783621135-1tecTULF0jOAjdf34lOl"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783621135-OZOH7tPhzmEAwKfqRABD"
upstream_id
"resp_04c91eba4aa8be6c016a4fe610d4188195b70c5a60a36bb1b1"
total_cost
0.03132
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
451
46.86%
Dialogue to Total Word Ratio
Ratio: 73.89%, Deviation: 3.89%
neededClean
false
wordsTotal
452
wordsDialogue
334
23.4281%