NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCR9N0SAMPF8DD48TJGCKJ5
Run Details
Model:
Claude Opus 5 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
22.0s
Cost:
$0.029965
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
81.27%
Matches word count
(details)
2.39%
Dialogue to Total Word Ratio
(details)
41.8322%
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 flat with the side of her hand. "Here's the perimeter," she said. "Bouncy castle, west. Trestle table with the sandwiches, north. Presents stacked here, by the patio doors." Bill squinted at the napkin. "That a dog?" "That's the gazebo." "It's got a tail." "It's got a *drainpipe*, Bill. Focus." He picked up his coffee, then put it down without drinking any, which was a thing he did when he was nervous, and he'd been doing it for eleven minutes. "I just think," he said, "and I say this with love, that robbing a seven-year-old is beneath us." "We robbed a hospice." "We robbed a hospice *fundraiser*, which is different, and I've made my peace with it, mostly, in the small hours. But this is *Tilly Ferguson's birthday party*. There'll be a magician." "There'll be forty presents in unopened wrapping paper, and Grandad Ferguson gives cash in cards. Envelopes, Bill. Untraceable." "How do you know that?" "Because I babysat her in 2019 and the man has no imagination." Bill sat back. "So you know the family." "Which is precisely why I'll be wearing the head." "The head." "The costume, Bill. Full-body. There's a company in Dagenham does licensed characters for hire, forty quid an hour, and I've booked us as Sparkle Pony and Captain Bananas." "Which am I?" "Which do you think?" He rubbed his face with both hands. "I'm not being a banana." "You're being a banana, and you'll be *charming*, because Captain Bananas leads the conga line at three fifteen, which takes every child, every parent, and every phone camera into the front garden for approximately four minutes." "Four minutes." "Four minutes, in which Sparkle Pony carries the gift sack out the patio doors, over the fence, and into the van, because Sparkle Pony *comes with a sack*, Bill. That's her whole thing. She distributes friendship." "And nobody thinks it's odd she's leaving with it full?" "Nobody looks at the sack. Everybody looks at the head." Molly tapped the napkin twice. "It's the greatest weakness of the British public. Put a giant face on something and they stop asking questions." Bill was quiet for a while. Outside, a bus wheezed past. "What about the dad?" he said. "What about him?" "Big lad. Rugby neck. He came to the door when you were casing the wheelie bins." "He came to the door because you were *whistling*." "I whistle when I'm working." "Then you'll be pleased to know the plan accounts for you. At three ten you tell him the magician's van's got a flat and he'll be out on that pavement with a jack for a quarter of an hour, sweating and heroic and utterly delighted with himself." "That's cruel." "That's *hospitality*." She folded the napkin and slid it into her coat. "Now. Are you in, or do I have to do the conga alone?" Bill exhaled slowly, and finally drank his coffee. "I'm not doing the conga in a banana." "You'll be magnificent," Molly said, and signalled for the bill.
created_at
"2026-07-25T13:44:06.945Z"
7/25/2026, 1:44:06 PM
model
"anthropic/claude-opus-5-20260723"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3224
moderation_latency
(null)
generation_time
21800
tokens_prompt
75
tokens_completion
818
native_tokens_prompt
93
native_tokens_completion
1180
native_tokens_completion_images
(null)
native_tokens_reasoning
54
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
0.029965
router
(null)
provider_responses
0
endpoint_id
"76cb4608-f48c-483d-8da8-9957fb44244e"
id
"msg_011CdNsirCB753pyaqhC2qoD"
is_byok
false
latency
1536
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-1784987046-SO1DVkzzFeLeSEaeviD4"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784987046-9deT8Sdbfb1gdveu7ln5"
upstream_id
"msg_011CdNsirCB753pyaqhC2qoD"
total_cost
0.029965
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
81.27%
Matches word count
n/a
neededClean
false
words
512
2.39%
Dialogue to Total Word Ratio
Ratio: 78.64%, Deviation: 8.64%
neededClean
false
wordsTotal
515
wordsDialogue
405
41.8322%