NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-06-01-01KT1GNNHFXYXB0A05BK02ETVD
Run Details
Model:
Claude Opus 4.8 (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
22.6s
Cost:
$0.028115
Provider:
Anthropic
Evaluation summary
Score
Evaluator
Details Link
99.98%
Matches word count
(details)
33.87%
Dialogue to Total Word Ratio
(details)
66.9254%
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 blueprints across the hood of the rusted sedan, smoothing the crinkled paper with the reverence of a surgeon prepping for an operation. The afternoon sun glinted off her sunglasses, which she absolutely did not need at 4 PM in a residential cul-de-sac. Bill leaned over, squinting at the diagram. It was, upon closer inspection, a hand-drawn map of a backyard with a bouncy castle labeled "DANGER ZONE" in red marker. "You drew this?" he asked. "Eight hours of reconnaissance," Molly said. "I watched that house like a hawk." "You fell asleep in the bushes." She waved a dismissive hand. "Strategic resting. The point is, I know everything. Cake at two, presents at three, parents distracted by the magician at three-fifteen." Bill scratched his stubble, knocking loose a crumb from his earlier surveillance donut. The presents were stacked on a folding table, wrapped in glittering paper and topped with bows the size of dinner plates. Behind them, a dozen children in party hats orbited a piƱata like tiny, sugar-fueled planets. "I dunno, Molly. Robbing a seven-year-old feels like a new low. Even for us." "It's not stealing if it's redistribution," she said. "Think of all those presents. Some kid's gonna have, like, forty toys. That's greedy. We're teaching a valuable lesson about materialism." "By taking the materials." "Exactly. You're getting it." Their plan was elegant in its stupidity. Bill would create a diversion. Molly would slip behind the snack table, load the gifts into a garbage bag, and vanish over the back fence before anyone noticed. The hard part, they both agreed, was the clown. "What's the diversion?" Bill whispered, though no one was within thirty feet of them. Molly handed him a wad of orange fabric. He unfolded it slowly, his face curdling into despair. "No." "It's the only way in." "I am not dressing as a clown." "You said you'd do anything for a score." "I said anything *reasonable*. Wigs are not reasonable." She clapped him on the shoulder, hard enough to stagger him. "Bill, the children won't suspect a thing. You'll blend in. Honk the nose, juggle a little, draw their eyes. Meanwhile, I work the magic." He stared at the rainbow wig in his hands like it had personally betrayed him. Across the lawn, a mother carried out a tray of cupcakes, and the children erupted into shrieks that could shatter glass. "This is humiliating." "This is professionalism." Forty minutes later, Bill stood at the edge of the party in a clown suit two sizes too small, sweating through the polyester while a child named Tyler relentlessly demanded a balloon animal. Molly crouched behind the table, garbage bag at the ready, watching the magician pull a sad rabbit from a hat. Then a small girl tugged Molly's sleeve. "Are you stealing the presents?" Molly froze. "What? No. I'm the, uh, present inspector." The girl considered this, then nodded gravely and walked away. Molly exhaled. "Children," she muttered, "are the easiest marks alive."
created_at
"2026-06-01T11:56:26.103Z"
6/1/2026, 11:56:26 AM
model
"anthropic/claude-4.8-opus-20260528"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1125
moderation_latency
(null)
generation_time
22510
tokens_prompt
77
tokens_completion
797
native_tokens_prompt
93
native_tokens_completion
1106
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
0.028115
router
(null)
provider_responses
0
endpoint_id
"b0d3beaa-5ff7-4666-a749-b1bb2b14c916"
id
"msg_01L7JnVeMxJV5iukeide2ph4"
is_byok
false
latency
1125
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-1780314986-ZXoetfDy9w3wSEWofq5O"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1780314986-C6edCk2ZElCDN4WhFvUw"
upstream_id
"msg_01L7JnVeMxJV5iukeide2ph4"
total_cost
0.028115
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Anthropic"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
99.98%
Matches word count
n/a
neededClean
false
words
498
33.87%
Dialogue to Total Word Ratio
Ratio: 34.65%, Deviation: 4.65%
neededClean
false
wordsTotal
505
wordsDialogue
175
66.9254%