NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCPWNTWNWSQRFND4J0MBNBV
Run Details
Model:
Claude Opus 5
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
21.6s
Cost:
$0.028290
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
90.48%
Matches word count
(details)
0.44%
Dialogue to Total Word Ratio
(details)
45.4598%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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 napkin across the diner table and smoothed it with the side of her hand. "Right," she said. "Blueprints." "That's a napkin." "It's a floor plan of the Hendersons' back garden, drawn to scale, and I'd appreciate a little professionalism." Bill leaned over it. "What's the wobbly bit?" "Bouncy castle." "And this?" "Also bouncy castle. There's two. It's a very indulgent household." She tapped a smear of ketchup. "Present table. Right there, under the gazebo. Twenty-two invitations went out, which means twenty-two gifts, minus the one from the divorced dad who always turns up with a garage voucher." "Twenty-one gifts." "Twenty-one gifts, Bill. Do you know what's in a seven-year-old's gift pile these days? Tablets. Headphones. Those little robot dogs that cost ninety quid and bite." "They bite?" "They nip. Focus." Bill sat back and folded his arms in the way he did when he was about to be difficult. "I want it on the record that I've got reservations." "Noted." "Moral reservations." "Also noted, filed, and set on fire. Go on then." "They're children, Moll. Small ones. With the little shoes." "They're consumers," said Molly, "who have done nothing to earn a robot dog. When I was seven I got a satsuma and a warning." "That's not their fault." "No, it's their grandparents' fault, and I intend to take it up with the grandparents by removing the evidence." She turned the napkin round. "Now. Access. Side gate, here, held shut with a bit of string because Gary Henderson thinks he's handy." "And the entertainment?" "Magician. Two till three. Which is when every single adult will be filming a man pull a scarf out of a rabbit and nobody will be watching a table." "So we walk in during the magic show and walk out with a wheelbarrow of Lego." "Bin bags. Wheelbarrows squeak." Bill rubbed his face. "What if a kid sees us?" "Then you'll be a clown." "I'll be what?" "There's a costume in the boot. Big shoes, red nose, terrible wig. If a child sees you, you honk the horn and do a little dance and they will absolutely not tell anyone, because children think clowns are supposed to be there and also slightly evil." "I hate clowns." "Everyone hates clowns, that's the beauty of it. Nobody looks directly at a clown. It's the closest thing this country has to invisibility." "And what are you doing while I'm honking?" "Bags. Boot. Gone by ten past." Bill stared at the napkin for a long moment. "What about the birthday girl?" "What about her?" "She'll cry." Molly picked up her tea and drank half of it. "Fine. We leave one." "One?" "One present, chosen at random, so she still gets a moment. That's my final offer and it's frankly ruinous." "Make it two." "Bill—" "Two, or I'm not being a clown." Molly looked at him. Bill looked back, jaw set, a man prepared to die on this particular hill in oversized shoes. "Two," she said. "But if either of them is the robot dog, I'm keeping it."
created_at
"2026-07-25T13:19:33.219Z"
7/25/2026, 1:19:33 PM
model
"anthropic/claude-opus-5-20260723"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1729
moderation_latency
(null)
generation_time
21160
tokens_prompt
75
tokens_completion
761
native_tokens_prompt
93
native_tokens_completion
1113
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
0.02829
router
(null)
provider_responses
0
endpoint_id
"76cb4608-f48c-483d-8da8-9957fb44244e"
id
"msg_011CdNqrC3ygi1M77j9TeQXF"
is_byok
false
latency
1729
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-1784985573-FISAD46ZEzbWbMA2gXZr"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784985573-5vLlJL0kHRYNydSu9XWh"
upstream_id
"msg_011CdNqrC3ygi1M77j9TeQXF"
total_cost
0.02829
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
90.48%
Matches word count
n/a
neededClean
false
words
510
0.44%
Dialogue to Total Word Ratio
Ratio: 80.43%, Deviation: 10.43%
neededClean
false
wordsTotal
516
wordsDialogue
415
45.4598%