NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-02-08-01KGYYB8PEMB45D1F1N3JKTTEV
Run Details
Model:
Claude Opus 4.6
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
19.0s
Cost:
$0.019250
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
99.74%
Matches word count
(details)
85.37%
Dialogue to Total Word Ratio
(details)
92.5569%
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
Molly pressed the binoculars against the windshield and squinted at the suburban house across the street. Balloons in every conceivable shade of pink bobbed from the mailbox, the porch railing, and what appeared to be a very unhappy golden retriever. "The target is a six-year-old named Penelope," she said, sliding a manila folder across the center console. "She's turning seven. Her parents are loaded. We're talking high-end presents. Possibly a pony." Bill opened the folder and stared at a photograph of a gap-toothed girl in a tutu. He closed it slowly. "We're robbing a child." "We're robbing a child's party. There's a distinction." "Is there, Molly? Is there really?" "Yes. Because the child will still have a birthday. She just won't have the mountain of consumer goods that accompanies it. If anything, we're teaching her a valuable lesson about materialism." Bill rubbed his temples. A clown car pulled into the driveway across the street. An actual clown car. Four clowns emerged from it in a honking, tumbling cascade of oversized shoes and seltzer bottles. Bill watched them with the hollow gaze of a man questioning every decision that had led him to this moment. "Walk me through the plan again," he said. "Glad you asked." Molly unrolled a hand-drawn blueprint across the dashboard. It was meticulous, color-coded, and smelled faintly of crayon. "The presents will be staged on a table in the backyard, right next to the bounce house. At exactly two fifteen, the clowns begin their act in the front yard. Every child and parent moves to watch. That gives us a four-minute window." "Four minutes to steal presents from a birthday party while clowns perform a distraction they don't know they're performing." "Exactly. We go in through the side gate, load the presents into these." She held up two enormous canvas bags printed with cartoon reindeer. "If anyone spots us, we say we're from the party supply company doing a gift-wrapping upgrade." "Nobody is going to believe that." "Bill, these people hired four clowns and a petting zoo for a seven-year-old. They will believe anything that sounds like a premium service." He couldn't argue with that. Through the binoculars he watched a man in khaki shorts carry an enormous box wrapped in glittering paper toward the backyard. The box was nearly as tall as he was. "What do you think that one is?" Bill asked. "That's the pony." "You can't wrap a pony." "Rich people can wrap anything." A miniature train emerged from behind the house, carrying squealing children along a track that circled the property. Molly's eyes went wide with genuine admiration. "These people really committed to the theme," she whispered. "What is the theme?" "Excess." Bill sighed and pulled a black ski mask from the glove compartment. He looked at it, then at the balloons, then at Molly. "If we get caught, I'm telling them you planned everything." "Obviously. Now put on your reindeer bag and let's go steal from a child."
created_at
"2026-02-08T15:36:44.850699+00:00"
2/8/2026, 3:36:44 PM
model
"anthropic/claude-4.6-opus-20260205"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1586
moderation_latency
181
generation_time
18482
tokens_prompt
61
tokens_completion
652
native_tokens_prompt
75
native_tokens_completion
755
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)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.01925
router
(null)
provider_responses
0
id
"61d618e5-a27f-4e1f-b16b-2174b9813eb1"
status
200
is_byok
false
latency
1585
endpoint_id
"a8708fb9-08ea-43bb-a3b1-165c87ef09df"
provider_name
"Amazon Bedrock"
model_permaslug
"anthropic/claude-4.6-opus-20260205"
api_type
"completions"
id
"gen-1770564985-lWM8nDFVZ7hS755U0Gs7"
upstream_id
"61d618e5-a27f-4e1f-b16b-2174b9813eb1"
total_cost
0.01925
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
Evaluation details
Result
Evaluator
Details
Meta Data
99.74%
Matches word count
n/a
neededClean
false
words
496
85.37%
Dialogue to Total Word Ratio
Ratio: 51.78%, Deviation: 1.78%
neededClean
false
wordsTotal
506
wordsDialogue
262
92.5569%