NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-05-31-01KSYXMJ51Y21GXAAA3T6CWF3C
Run Details
Model:
Claude Opus 4.8 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
20.0s
Cost:
$0.027690
Provider:
Anthropic
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
49.9995%
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 table, smoothing out the wrinkles with the reverence of a surgeon preparing for a delicate operation. The document in question was a child's crayon drawing of a backyard, complete with a lopsided sun wearing sunglasses. "This is what we're working with?" Bill leaned over, squinting at the paper. "Did you steal this from a refrigerator?" "I liberated it from intelligence sources," Molly said, jabbing a finger at a green scribble. "The bounce house is here. Maximum chaos zone. Children will be everywhere, screaming, sticky, unpredictable." Bill pulled out a notepad, his face grave. "And the presents?" "The gift table." She tapped a rectangle near the swing set. "Stacked high. Birthday haul for a seven-year-old. We're talking premium loot, Bill. The kind of toys that haven't even been opened yet. Still in the plastic. Pristine." Bill let out a low whistle. He'd been in this business for fifteen years, and nothing made his heart race like the words *still in the plastic*. "Security?" Molly's eyes darkened. "Worse than you'd think. We've got a clown." The notepad slipped from Bill's fingers. He retrieved it slowly, his hands trembling. "A clown? You didn't say anything about a clown." "His name is Mr. Giggles. Word on the street is he does balloon animals so realistic, they bite." She slid a grainy photograph across the table. A man in oversized shoes grinned maniacally beside a folding table of juice boxes. "And there's a magician booked for two o'clock. We hit before he arrives, or all bets are off. Last thing we need is some guy pulling our getaway car out of a hat." Bill studied the photo, then loosened his collar. "What's the escape route?" "Through the hedge, past the dog, and over the fence." Molly traced the path with her pinkie. "The dog's a golden retriever named Biscuit. Friendly, but easily distracted by tennis balls, which I'll be carrying as a tactical countermeasure." "And the parents?" "That's the beautiful part." Molly leaned back, lacing her fingers together with the satisfaction of a woman who had thought of everything. "Parents at these parties are exhausted. Sleep-deprived. They won't notice two extra adults if we look like we belong. You'll be Uncle Bill. I'll be Aunt Molly. We blend in, we grab the goods during the cake-cutting distraction, and we vanish." Bill frowned. "What if someone asks whose side of the family we're on?" "Then you cry," Molly said simply. "Nobody questions a crying adult at a child's party. It's social kryptonite." He nodded slowly, absorbing the wisdom. After a long moment, he set down his pen. "Molly, in all my years, I've cracked vaults, dodged lasers, and outrun the law on three continents. But a children's party? This is the big leagues." "That's why I called you." She stood, extending her hand. "Are you in?" Bill rose, gripping it firmly. "For the loot still in the plastic? I'm in." "Excellent." Molly grinned. "Now, can you fit into a party hat?"
created_at
"2026-05-31T11:45:17.997Z"
5/31/2026, 11:45:17 AM
model
"anthropic/claude-4.8-opus-20260528"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1047
moderation_latency
(null)
generation_time
19948
tokens_prompt
77
tokens_completion
779
native_tokens_prompt
93
native_tokens_completion
1089
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
0.02769
router
(null)
provider_responses
0
endpoint_id
"b0d3beaa-5ff7-4666-a749-b1bb2b14c916"
id
"msg_01J9JVHckyw8aK9F5DSHbyJt"
is_byok
false
latency
1047
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-1780227917-CAleI3YYbWjfLvD4pRD6"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1780227917-FbuR3VMXegfguDwcTymN"
upstream_id
"msg_01J9JVHckyw8aK9F5DSHbyJt"
total_cost
0.02769
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
501
0.00%
Dialogue to Total Word Ratio
Ratio: 59.80%, Deviation: 29.80%
neededClean
false
wordsTotal
505
wordsDialogue
302
49.9995%