NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-01-01KWE9TQPHPS0DNM671DQXS85H
Run Details
Model:
Claude Sonnet 5
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
16.7s
Cost:
$0.011126
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
68.10%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
34.0511%
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 warehouse smelled like stale popcorn and broken dreams, which Bill found appropriate given what they were about to do. He spread the blueprint across the table—a hand-drawn map of the Kowalski residence, complete with a suspiciously detailed rendering of a bouncy castle. "It's a six-year-old's birthday party, Bill," Molly said, arms crossed. "You drew the bouncy castle to scale." "Details matter, Molly. Details are the difference between a clean job and a life sentence." "We're stealing birthday presents from children." "We're liberating overpriced plastic garbage from a system that teaches kids materialism before they can spell their own names." Bill tapped the map with a pen. "Also, some of those presents are worth actual money. Little Tommy's getting a gaming console from his uncle. Cash value, two hundred and fifty dollars." Molly rubbed her temples. "There will be forty parents there. Forty parents with phones. Forty parents who will absolutely call the police the second they see two adults in ski masks running off with a pile of Paw Patrol merchandise." "Which is why we're not wearing ski masks." Bill grinned and reached into his bag, pulling out two folded costumes. "We're wearing these." He held them up. Two mascot costumes—a purple dinosaur and a pink unicorn, both slightly moth-eaten, both purchased from a defunct children's entertainment company that had gone bankrupt years ago. "You want me to steal presents dressed as a unicorn." "I want you to *blend in*, Molly. Nobody suspects the unicorn. The unicorn is doing balloon animals. The unicorn is beloved." "The unicorn is going to get present duty while you do balloon animals?" "The unicorn is going to distract the children while the dinosaur—that's me—loads the loot bag into the getaway vehicle." Bill said this with the confidence of a man who had clearly rehearsed it in a mirror. "It's elegant. It's foolproof." "Nothing with you is foolproof. Remember the jewelry store? You got your dinosaur tail—wait, you didn't have a tail then. You got your regular human foot stuck in the vent." "That was a structural issue, not a planning issue." Molly picked up the unicorn head and stared into its vacant, painted eyes. "If I get caught wearing this, I want it on record that this was your idea." "Duly noted. Now—" Bill unfolded a second sheet of paper, this one a timeline scrawled in red marker. "Party starts at two. Cake at three. Presents at three-thirty. We arrive at three-fifteen, you do your unicorn thing, I locate the present table, we're out by three-forty before anyone notices the loot bag is actually a loot bag." "And if a kid asks the unicorn a direct question?" "Improvise. You're a professional. Also unicorns can't talk, it's a known fact, everyone knows this, we're covered." Molly sighed and pulled the unicorn head down over her own, her voice muffling instantly into something distant and doomed. "If I end up on a milk carton, Bill, I'm haunting you specifically." "Noted," Bill said, already zipping himself into purple felt. "Let's go steal some birthday presents."
created_at
"2026-07-01T07:38:23.324Z"
7/1/2026, 7:38:23 AM
model
"anthropic/claude-sonnet-5-20260630"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1526
moderation_latency
(null)
generation_time
16666
tokens_prompt
77
tokens_completion
809
native_tokens_prompt
93
native_tokens_completion
1094
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
0.011126
router
(null)
provider_responses
0
endpoint_id
"37b16f45-eeae-45f3-9b85-2137f7c6d61b"
id
"msg_016Q8W7P454KZqJJMeDuYFzU"
is_byok
false
latency
1463
model_permaslug
"anthropic/claude-sonnet-5-20260630"
provider_name
"Amazon Bedrock"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.17.0; linux; x64))"
http_referer
(null)
request_id
"req-1782891503-v9ItR9MBhZfWDcBm61qG"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1782891503-webfSSwmsnux54htI5ps"
upstream_id
"msg_016Q8W7P454KZqJJMeDuYFzU"
total_cost
0.011126
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
68.10%
Matches word count
n/a
neededClean
false
words
514
0.00%
Dialogue to Total Word Ratio
Ratio: 66.03%, Deviation: 36.03%
neededClean
false
wordsTotal
521
wordsDialogue
344
34.0511%