NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCR6XF7T6GEVJM26ZC6C0CG
Run Details
Model:
Claude Opus 5 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
21.8s
Cost:
$0.029915
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
68.10%
Matches word count
(details)
0.01%
Dialogue to Total Word Ratio
(details)
34.0538%
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 van smelled of cold pizza and ambition. Molly had spread the blueprints across the dashboard — technically a hand-drawn map on the back of a takeaway menu, but she'd used three colours of pen, which made it a blueprint in spirit. "Right," she said. "Bouncy castle here. Trestle table of presents here. Two exits, one nan." Bill squinted at the menu. "Why is the nan circled in red?" "Because the nan is the problem, Bill. Nans see everything. Nans have a sixth sense specifically calibrated to detect people being where they shouldn't." "So we distract the nan." "We *neutralise* the nan." Molly tapped the paper. "Sherry. There's always a bottle of cream sherry at these things, and there's always a nan who considers it her civic duty to finish it. We simply move it closer to her." Bill turned this over with the slow gravity of a man who had once been described by a magistrate as "not the ringleader." "That's actually clever." "I have my moments." She folded the menu and shoved it in the glovebox, where it joined nine other schemes of varying success. "Now. Your job is the entrance. There'll be a woman with a clipboard checking names." "At a seventh birthday party?" "It's Tunbridge Wells, Bill. There's a guest list, an allergy register, and probably a risk assessment for the piñata." "So what do I say?" "You say you're the entertainer." Bill went very still. "I'm not the entertainer." "You're wearing a purple waistcoat." "That's my funeral waistcoat!" "And to a seven-year-old, that's magic." Molly reached into the back and produced a plastic bag containing three balloons and a wilting bunch of fake flowers. "Congratulations. You're Bill the Bewildering." "I can't do magic." "Nobody at a children's party can do magic. That's the whole industry." She handed him the bag. "Make a balloon animal. Any animal. Confidence is ninety per cent of it." Bill inflated a balloon, twisted it twice, and produced something that looked like a diagram of a hernia. "What's that?" "Giraffe." "It's got no neck, Bill. A giraffe is *mostly* neck." "It's crouching." Molly closed her eyes for a moment, the way saints do in paintings. "Fine. It's abstract. Children love abstract. Meanwhile I'll be in the kitchen, because the presents always migrate to the kitchen when the cake comes out, and everyone's looking at the cake. Nobody guards a gift table during 'Happy Birthday.'" "And then we just... take them?" "We take them out through the side gate, into the van, and we're on the A21 before anyone's blown out the candles." She paused. "Except one." "Except one?" "We leave one. The biggest one. The kid gets a present, we get thirty, and nobody has to watch a child cry, which frankly I'm not built for." Bill nodded, oddly moved. "That's the most decent thing you've ever said." "Don't tell anyone." Molly checked her watch and started the engine. "Party's at two. Balloon practice till then." Bill squeaked out another twist. The balloon burst instantly, and somewhere in Molly's chest, a small alarm began quietly ringing.
created_at
"2026-07-25T13:42:37.296Z"
7/25/2026, 1:42:37 PM
model
"anthropic/claude-opus-5-20260723"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
6690
moderation_latency
200
generation_time
21521
tokens_prompt
75
tokens_completion
844
native_tokens_prompt
93
native_tokens_completion
1178
native_tokens_completion_images
(null)
native_tokens_reasoning
60
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
0.029915
router
(null)
provider_responses
0
endpoint_id
"76cb4608-f48c-483d-8da8-9957fb44244e"
id
"msg_011CdNscEvDRDjeqhF5YdSno"
is_byok
false
latency
1665
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-1784986957-kVkERGIrcqn4RFmxs8Wk"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784986957-CZzHKO7SDZIMCSEPWt8q"
upstream_id
"msg_011CdNscEvDRDjeqhF5YdSno"
total_cost
0.029915
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
0.01%
Dialogue to Total Word Ratio
Ratio: 64.02%, Deviation: 14.02%
neededClean
false
wordsTotal
517
wordsDialogue
331
34.0538%