NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-01-01KWE9H98QFT1BG7CYN8TM9WXE
Run Details
Model:
Claude Sonnet 5
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
17.4s
Cost:
$0.010716
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
99.92%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
49.9595%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% 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 slid the blueprint across the sticky diner table, narrowly avoiding a puddle of syrup, and jabbed her finger at a crudely drawn rectangle labeled "BOUNCY CASTLE." "This is our way in," she said. Bill squinted at it, turning his head sideways like that would somehow make the drawing more sophisticated. "That's a bouncy castle, Molly. You want us to break into a seven-year-old's birthday party through a bouncy castle." "I want us to break into a seven-year-old's birthday party through a bouncy castle, yes." "We rob banks, Molly. We robbed the Second National Trust. We robbed an armored truck on the interstate. And now you're telling me our next job is a kid named Tyler's Paw Patrol-themed birthday bash?" "Do you know how much loot is at these things now?" Molly leaned forward, eyes gleaming with the particular fire she got before a big score. "Bill, I did recon. Tyler's getting a drone, two gaming consoles, gift cards—actual gift cards, Bill, liquid assets—and I heard his aunt got him a watch. A real one. Gold-plated." Bill rubbed his temples. "A gold-plated watch for a seven-year-old." "Rich people are insane, Bill, that's not new information." "So what's the plan? We just walk in, dressed as clowns, and steal a mountain of Nerf guns?" "Better." Molly pulled out a second piece of paper—this one a costume sketch, if you were feeling generous with the word "sketch." "You're the balloon guy. I'm the face painter. We get invited in legitimately, we set up shop, and while thirty screaming children are distracted by tiger stripes and dachshund balloon animals, we load the presents into the panel van parked out back." "I don't know how to make balloon animals." "You don't need to know how to make balloon animals, Bill, you need to know how to make a balloon look vaguely like a dog while I empty a gift table into a duffel bag." "And when the kids notice their presents are missing?" "They won't. Not for a while. Kids at parties don't track inventory, Bill, they're eating cake and hopped up on Hi-C. By the time anyone realizes Tyler's drone is gone, we're four exits down the highway." Bill sat back, arms crossed, clearly wrestling with something deeper than logistics. "There's a line, Molly. There's a line we don't cross, and I feel like a kid's birthday party might be past that line." "It's not the kid's money, Bill. It's the parents'. Insured, overpriced, guilt-gift money. We're basically doing social commentary." "That's not a defense, that's a TED talk." "Look, you in or not? Because Tyler's party is Saturday at two, and if we don't case the bounce house today, someone else might beat us to a booth full of unsupervised gift bags." Bill exhaled, long and slow, already reaching for the balloon sketch. "Fine. But I'm not doing the voice. Balloon guys always do a voice, and I refuse." "Deal," Molly said, already grinning. "No voice."
created_at
"2026-07-01T07:33:13.632Z"
7/1/2026, 7:33:13 AM
model
"anthropic/claude-sonnet-5-20260630"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1712
moderation_latency
(null)
generation_time
17337
tokens_prompt
77
tokens_completion
756
native_tokens_prompt
93
native_tokens_completion
1053
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
0.010716
router
(null)
provider_responses
0
endpoint_id
"37b16f45-eeae-45f3-9b85-2137f7c6d61b"
id
"msg_01QWCz5S1JKCjXrwQjHCThoz"
is_byok
false
latency
1712
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-1782891193-vrIpd4GXjmOUhr5yfjzY"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1782891193-Iy6mTZHmbpTyEHnbRaI5"
upstream_id
"msg_01QWCz5S1JKCjXrwQjHCThoz"
total_cost
0.010716
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
99.92%
Matches word count
n/a
neededClean
false
words
497
0.00%
Dialogue to Total Word Ratio
Ratio: 78.15%, Deviation: 28.15%
neededClean
false
wordsTotal
508
wordsDialogue
397
49.9595%