NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-06-01-01KT1GHRPR8MD3QQ28M3Z5BSR0
Run Details
Model:
Claude Opus 4.8 (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
21.5s
Cost:
$0.026690
Provider:
Anthropic
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
3.01%
Dialogue to Total Word Ratio
(details)
51.5059%
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 blueprint across the hood of the rusted Camry, weighing down the corners with empty energy drink cans. The fluorescent glow of the convenience store sign painted her face an unnatural shade of orange. "It's a fortress," she said, tapping the paper. "Bouncy castle here. Petting zoo here. And the presents—" She jabbed her finger triumphantly. "Stacked on a folding table guarded by nothing but a clown." Bill leaned in, squinting through his reading glasses. He'd insisted on the disguise: a polo shirt and khakis that made him look like a substitute gym teacher. "What kind of clown?" "Does it matter?" "It matters, Molly. There's a big difference between a balloon-animal clown and a face-painting clown. One of them has scissors." She rubbed her temples. They'd been at this for three hours, and Bill's commitment to clown taxonomy was eroding her will to live. The plan was simple. It was always simple. They'd blend in among the parents, locate the gift table, and exfiltrate the loot during the inevitable chaos of the piñata. Children swinging blindfolded bats created natural confusion. It was, Molly believed, the perfect cover. "The birthday boy is turning seven," she continued. "His name is Tyler. We arrive at two, we leave by two-fifteen, we're millionaires." "Millionaires." Bill snorted. "From a kid's birthday party." "You'd be amazed what these suburban parents spend. I'm talking gaming consoles. Drones. One kid last week got a pony." "We can't fit a pony in the Camry." "We're not stealing the pony, Bill." He removed his glasses and polished them on his shirt, a gesture he performed whenever he was about to say something he considered profound. Molly braced herself. "Maybe," he said slowly, "this is beneath us. We used to do museums. Banks. Remember the casino?" "The casino where you got bitten by a service dog?" "That dog had it out for me." Molly folded the blueprint with the precision of a woman who had folded many incriminating documents. The truth, which she would never admit to Bill, was that the museums had stopped calling. The banks had better cameras now. And last month, when she'd tried to pawn a stolen Rolex, the clerk had laughed at her. A children's party was a step down, yes, but it was also, crucially, a step they could actually accomplish. "Look," she said. "We grab the presents, we hit the next party two towns over, and within a month we've got enough to disappear. Somewhere warm. Somewhere with no extradition and excellent guacamole." Bill considered this. A garbage truck rumbled past, scattering pigeons. "What about the cake?" "What about it?" "Do we take the cake?" Molly stared at him for a long moment, searching his face for any trace of irony and finding, as always, none. "Sure, Bill," she sighed, climbing into the driver's seat. "We take the cake." He grinned, sliding in beside her, and buckled his seatbelt with the satisfaction of a man whose only demand had finally been met.
created_at
"2026-06-01T11:54:18.206Z"
6/1/2026, 11:54:18 AM
model
"anthropic/claude-4.8-opus-20260528"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3442
moderation_latency
(null)
generation_time
21510
tokens_prompt
77
tokens_completion
785
native_tokens_prompt
93
native_tokens_completion
1049
native_tokens_completion_images
(null)
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
"end_turn"
service_tier
"default"
usage
0.02669
router
(null)
provider_responses
0
endpoint_id
"b0d3beaa-5ff7-4666-a749-b1bb2b14c916"
id
"msg_01VjDzgkhMJdjxRKSwoNzwow"
is_byok
false
latency
3131
model_permaslug
"anthropic/claude-4.8-opus-20260528"
provider_name
"Anthropic"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.16.0; linux; x64))"
http_referer
(null)
request_id
"req-1780314858-yk7mDB50D3UWCQOa92dG"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1780314858-uJjEpUGNJdBfO1c14WuA"
upstream_id
"msg_01VjDzgkhMJdjxRKSwoNzwow"
total_cost
0.02669
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Anthropic"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
3.01%
Dialogue to Total Word Ratio
Ratio: 38.37%, Deviation: 8.37%
neededClean
false
wordsTotal
503
wordsDialogue
193
51.5059%