NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCR3HTBKR87XWT5JK7KVPC1
Run Details
Model:
Claude Opus 5 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
23.2s
Cost:
$0.030915
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
98.71%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
49.3562%
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
The blueprints were drawn on the back of a pizza box, which Bill considered a professional touch. "Bouncy castle here," he said, tapping grease-stained cardboard with a pencil. "Trestle table of sandwiches here. And the presents, Molly, the presents are stacked in the conservatory like ingots in a vault." "How many?" "Forty-one. I counted them through the hedge while pretending to be a man looking for his dog." "Do you have a dog?" "I have a lead," said Bill. "A lead implies a dog. That's misdirection." Molly leaned back in the deckchair and considered the ceiling of the lock-up, where a strip light flickered with what she felt was appropriate menace. She had done banks. She had done a jeweller's in Antwerp with a man named Klaus who wept during the getaway. She had never done a seventh birthday party. "Security?" "Fourteen children, four adults, one grandmother who I'm told does not sit down." "And the mother?" "Deborah." Bill said the name the way other men say *Interpol*. "Deborah has a laminated schedule. Games at two. Cake at half past. She will notice if a single sausage roll goes missing, never mind forty-one presents." "Then we go during the entertainer." "Explain." "Magician arrives at two-fifteen. Lights down, curtains shut, everyone facing the same direction for twenty-five minutes." Molly sat forward. "Every adult in that house will be standing at the back with a glass of warm Prosecco, thinking about their mortgage. Nobody looks at a conservatory during a magic show. Nobody looks at anything." Bill's pencil hovered. "That's beautiful." "It's basic crowd psychology." "It's *beautiful*, Molly, don't spoil it." He wrote MAGICIAN in capitals and underlined it twice. "What's my role?" "You're the uncle." "Which uncle?" "Every party has an uncle nobody can place. Stands by the crisps. Asks children how school is going. Nobody challenges the uncle because everyone assumes he belongs to somebody else." "I'll need a fleece." "You'll need a fleece and a card with a tenner in it, in case someone hands you the birthday girl." "And you?" "Caterer." Molly produced a hi-vis tabard and a large flat box. "I arrive at two-ten with an unrequested platter of vol-au-vents. Deborah will hate them, but she won't turn them away, because turning away food in front of other mothers is social suicide. That gets me a door propped open and a legitimate reason to be carrying a large empty box out of the house." Bill was quiet for a moment. "There's an ethical dimension here I want to raise." "Go on." "They're children's presents." "They're forty-one unopened items, average retail forty quid, of which thirty are duplicates and eleven are educational." Molly shrugged. "We're preventing disappointment. We're a service." "And the birthday girl?" "We leave her the pony one. I'm not a monster." She folded the tabard with a bank clerk's precision. "Two o'clock, Bill. Fleece. Crisps. Look bored." "I *am* bored," said Bill. "That's why I said yes." "Good," said Molly, switching off the flickering light. "Boredom's the best disguise there is."
created_at
"2026-07-25T13:40:47.317Z"
7/25/2026, 1:40:47 PM
model
"anthropic/claude-opus-5-20260723"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3920
moderation_latency
(null)
generation_time
22855
tokens_prompt
75
tokens_completion
845
native_tokens_prompt
93
native_tokens_completion
1218
native_tokens_completion_images
(null)
native_tokens_reasoning
53
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
0.030915
router
(null)
provider_responses
0
endpoint_id
"76cb4608-f48c-483d-8da8-9957fb44244e"
id
"msg_011CdNsU7hAjDDDLGYoAfTF8"
is_byok
false
latency
1654
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-1784986847-XCuwMOp07cnKlM9cQDWP"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784986847-f7l938xXIDE7FdTSJJT3"
upstream_id
"msg_011CdNsU7hAjDDDLGYoAfTF8"
total_cost
0.030915
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
98.71%
Matches word count
n/a
neededClean
false
words
506
0.00%
Dialogue to Total Word Ratio
Ratio: 71.95%, Deviation: 21.95%
neededClean
false
wordsTotal
517
wordsDialogue
372
49.3562%