NC Bench
Overview
About
Tests
Models
Vendors
Compare
Run: 2026-02-24-01KJ8TR4GEZXSCA3M7Y86182QA
Run Details
Model:
Claude Opus 4.6 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
11.9s
Cost:
$0.011260
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
95.99%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
97.9930%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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
"Alright, Bill, here's the layout. Bouncy castle on the north side, face painting station to the east, and the present table is dead center, guarded by a woman in a Elsa costume," Molly said, spreading a crayon-drawn map across the dashboard. "That's a lot of security." "It gets worse. There's a balloon artist near the back gate. He'll spot us immediately." "Can we bribe him?" "He's six, Bill. He works for juice boxes and animal crackers." "So we bring animal crackers." "We're not bribing a child." "You literally want to steal from children." "That's different. That's professional. Bribery is tacky." Bill rubbed his temples. "Fine. What's the entry plan?" "You go in disguised as a clown." "Absolutely not." "You already have the shoes for it." "Those are orthopedic, Molly." "Whatever. You waddle in, create a distraction, and I grab the presents during the chaos." "What kind of distraction?" "Make balloon animals." "I don't know how to make balloon animals." "Then just pop them and cry. Kids love watching adults fall apart. While everyone's staring at your meltdown, I load the van." Bill stared at her for a long moment. "This is our worst plan yet." "And yet you're going to say yes." "I need a bigger shoe budget."
created_at
"2026-02-24T22:01:39.093Z"
2/24/2026, 10:01:39 PM
model
"anthropic/claude-4.6-opus-20260205"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1580
moderation_latency
407
generation_time
11359
tokens_prompt
77
tokens_completion
418
native_tokens_prompt
92
native_tokens_completion
432
native_tokens_completion_images
(null)
native_tokens_reasoning
71
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
0.01126
router
(null)
provider_responses
0
endpoint_id
"a8708fb9-08ea-43bb-a3b1-165c87ef09df"
id
"f4e31370-5267-464b-87d1-a7c6e81c434a"
is_byok
false
latency
1545
model_permaslug
"anthropic/claude-4.6-opus-20260205"
provider_name
"Amazon Bedrock"
status
200
api_type
"completions"
id
"gen-1771970499-hdRSIy60tv8ovCe6leHB"
upstream_id
"f4e31370-5267-464b-87d1-a7c6e81c434a"
total_cost
0.01126
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
Evaluation details
Result
Evaluator
Details
Meta Data
95.99%
Matches word count
n/a
neededClean
false
words
208
100.00%
Dialogue to Total Word Ratio
Ratio: 89.47%, Deviation: 0.53%
neededClean
false
wordsTotal
209
wordsDialogue
187
97.9930%