NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCPV7ZK9ES39KKVMGPBSWAY
Run Details
Model:
Claude Opus 5
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
24.2s
Cost:
$0.026990
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
81.27%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
40.6382%
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
The van smelled of stale crisps and betrayal, which Molly maintained was Bill's natural cologne. "Right," she said, spreading a napkin across the dashboard. "Here's the layout. Bouncy castle, north lawn. Trestle table with the sandwiches, west. Presents, and this is critical, Bill, presents are stacked in the conservatory." "Why's that critical?" "Because the conservatory has glass walls, so everyone can see us doing crimes in it." "That does seem like a flaw in the plan." "It's a flaw in the *house*, Bill. I'm working with what the architect gave me." Bill leaned over the napkin. "What's this circle you've drawn here?" "That's the clown." "You've drawn him quite large." "He *is* quite large. His name is Mr. Bimbles and he has forearms like a stevedore. I watched him twist a balloon giraffe yesterday and I heard something crack." "The balloon?" "His wrist, I think. He didn't seem bothered." Bill sat back and folded his arms. "I want it on record that I said we should do the bank." "The bank has a vault." "The bank doesn't have Mr. Bimbles." "The bank," Molly said, "has a vault, six cameras, and a man named Derek who *wants* to be a hero, Bill. He's been waiting his whole life. He's got a look about him. Whereas this house has forty children, two exhausted parents, and roughly nine hundred pounds of unattended Lego." "Lego's worth something?" "Lego holds its value better than gold. Look it up. Now. You're going in as the entertainer." "I don't have an act." "You have a hat." "A hat isn't an act." "You'd be amazed how far a hat gets you with a six-year-old. Wear the hat, do the thing with the coin, keep them in the garden. I'll be in the conservatory with the bin bags." "How many bin bags?" "Twelve." "*Twelve?*" "It's a big party, Bill. The birthday girl's grandmother owns a chain of garden centres. There's a rocking horse in there I could sell to a museum." Bill was quiet for a moment. "And if someone comes in?" "Then I'm the wrapping lady." "The what?" "The wrapping lady. Hired to gather the presents so the child can open them dramatically, one at a time, on the stage." "There's a stage?" "There will be by then. I've told them there's a stage. They believed me because I said it in a clipboard voice." She demonstrated. "*Excuse me, I'll need those in the van for the reveal.*" "That's frighteningly good." "Thank you." "One thing." Bill tapped the napkin. "This says the party starts at two. It's ten past two." Molly looked at the dashboard clock. Then at the house, where a bouncy castle was inflating like a slow orange lung. "Right," she said. "Hat on." "I've lost the coin." "Then do the thing with your thumb." "Children hate the thing with my thumb." "Bill," Molly said, opening the door, "so does everyone. Go be beloved."
created_at
"2026-07-25T13:18:46.265Z"
7/25/2026, 1:18:46 PM
model
"anthropic/claude-opus-5-20260723"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1965
moderation_latency
(null)
generation_time
24113
tokens_prompt
75
tokens_completion
723
native_tokens_prompt
93
native_tokens_completion
1061
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
0.02699
router
(null)
provider_responses
0
endpoint_id
"76cb4608-f48c-483d-8da8-9957fb44244e"
id
"msg_011CdNqnkG7dKtSn5bFLuYBQ"
is_byok
false
latency
1965
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-1784985526-yopkJaywmyH8iXnVrP9V"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784985526-A4bgRTC3TEfsHpZnX46i"
upstream_id
"msg_011CdNqnkG7dKtSn5bFLuYBQ"
total_cost
0.02699
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
488
0.00%
Dialogue to Total Word Ratio
Ratio: 84.29%, Deviation: 14.29%
neededClean
false
wordsTotal
490
wordsDialogue
413
40.6382%