NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCPWV2J38XRF2RDX4Z1P9HT
Run Details
Model:
Claude Opus 5
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
22.6s
Cost:
$0.027115
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
9.61%
Matches word count
(details)
20.78%
Dialogue to Total Word Ratio
(details)
15.1918%
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
The van smelled of cold pizza and ambition. Molly had the blueprints spread across the dashboard, though calling them blueprints was generous — it was a napkin from a diner, marked in three colors of crayon she'd borrowed from a booster seat. "That's the bouncy castle," she said, tapping a purple blob. Bill leaned in, squinting. "I thought that was the cake." "The cake's the yellow one." "Why is the cake bigger than the house?" "Because it's a very serious cake, Bill. Six tiers. The father is a dentist with something to prove." Bill sat back and unwrapped a stick of gum with the solemnity of a man loading a firearm. Twenty-two years he'd been in the game. Armored cars. A yacht, once, briefly, before the tide came in. And now here he was, parked outside a cul-de-sac in Willow Bend, staking out the ninth birthday of a child named Kayleigh Pruitt. The house was a monument to excess. Balloon arches. A rented pony that looked like it had seen things. And in the sunroom, visible through the bay window, a pile of gift-wrapped boxes stacked in a glittering pyramid that would have made a pharaoh weep with envy. "Forty-one presents," Molly said, lowering the binoculars. "I counted twice. The grandmother alone brought six." "That's not a birthday. That's a hoard." "That's a *retirement plan*." She handed him the binoculars. "The tall one in the corner? That's a bicycle. And the flat one under the bear — that's a tablet, top of the line, still in the box, and there are seven of them because nobody at this party can communicate about anything." Bill made a small, reverent noise. They had rules, of course. Molly had insisted on rules, because she'd once dated a man who didn't have any and he was currently serving eleven years in Ohio. Rule one: nobody cries. If a child cried, they walked away, no argument, no discussion. Rule two: the pony stayed. Rule three — and this had taken an hour of negotiation over onion rings — Kayleigh got to keep one present, her choice, because Molly was a thief and not a monster, and there was a distinction, and she would defend it in court if she had to. "The window of opportunity is eleven minutes," she said. "Piñata to cake. Every adult in the backyard, every child holding a stick. Total anarchy. Nobody's looking at the sunroom." "Eleven minutes to move forty-one boxes." "Forty. She keeps one." "Forty boxes." "You've got a wagon." Bill turned to look at her, slowly, the way a man looks at a doctor who's just said something unforgivable. "A wagon." "A red one. It's in the back, under the tarp. I got it at a yard sale for four dollars and I want you to know I felt nothing." She checked her watch, then the house, then her watch again. A woman in a sunhat carried out a tray of juice boxes. Somewhere inside, a stereo began playing a song about sharing. Molly cracked her knuckles. "Balloon animals," she said. "You remember how?" Bill sighed and reached for the tarp. "I remember how."
created_at
"2026-07-25T13:19:38.583Z"
7/25/2026, 1:19:38 PM
model
"anthropic/claude-opus-5-20260723"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2655
moderation_latency
(null)
generation_time
22501
tokens_prompt
75
tokens_completion
758
native_tokens_prompt
93
native_tokens_completion
1066
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
0.027115
router
(null)
provider_responses
0
endpoint_id
"76cb4608-f48c-483d-8da8-9957fb44244e"
id
"msg_011CdNqrcVQUDwEbMz3g7wJY"
is_byok
false
latency
2653
model_permaslug
"anthropic/claude-opus-5-20260723"
provider_name
"Amazon Bedrock"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1784985578-9MD6eIeydaYr0WNoUm5s"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784985578-OqLdy7IYvl04D7083J49"
upstream_id
"msg_011CdNqrcVQUDwEbMz3g7wJY"
total_cost
0.027115
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
9.61%
Matches word count
n/a
neededClean
false
words
522
20.78%
Dialogue to Total Word Ratio
Ratio: 35.61%, Deviation: 5.61%
neededClean
false
wordsTotal
528
wordsDialogue
188
15.1918%