NC Bench
Overview
About
Tests
Models
Model
Claude Opus 4.6
Test
Dialogue tags
Scenario
Write 500 words with 70% dialogue
Run ID
2026-02-08-01KGYXTG8K7NMRM5KBJV1VP8Z3
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 hand-drawn map across the sticky diner table, knocking over the salt shaker with her elbow. She didn't bother picking it up. "Alright, here's the layout. The party starts at two. Bouncy castle in the back, face painting station by the garage, and the present table is right here, next to the sliding glass door." Bill squinted at the map, turning it sideways. "Why does that look like a horse?" "That's the bouncy castle, Bill." "You drew a horse for the bouncy castle." "It's a castle with turrets. Focus, please. We have a thirty-minute window between the arrival of the clown and the cutting of the cake. That's when every kid in that backyard will be hypnotized by balloon animals and we make our move." Bill leaned back in the booth and folded his arms. "I don't feel great about this one, Molly." "You didn't feel great about the nursing home bingo heist either, and we walked away with fourteen gift baskets and a case of butterscotch candies." "Those old ladies chased us for three blocks." "And they didn't catch us. That's the point. Now, you're going in through the side gate disguised as a pizza delivery guy. I called ahead and ordered six large pepperonis under the mom's name, so they'll be expecting you." "Wait, we're actually buying pizza for this?" "It's called an investment, Bill. You walk the pizzas to the kitchen, and while every parent swarms the food like seagulls on a french fry, I back the van up to the sliding glass door and load the presents." "What if a kid sees you?" "Kids don't care about presents until it's time to open them. They're feral little animals, Bill. They'll be smashing each other's faces into cake and screaming about whose turn it is on the bouncy castle. Trust me, I've done research. I watched eleven hours of birthday party videos online." "That's disturbing." "That's dedication. Now, the birthday boy's name is Aiden. He's turning seven, and according to his mom's social media, he asked for a gaming console, three different LEGO sets, and something called a Squishmallow. Do you know what those go for on resale?" "I honestly don't." "Enough to cover rent for two months. And that's just Aiden's haul. There'll be thirty kids at this party. Thirty kids means thirty presents, minimum. Some overachieving parents will bring two." Bill rubbed his temples slowly. "We're robbing children, Molly." "We're robbing a situation. The children are incidental. Besides, studies show kids are happier with fewer toys. We're practically doing a public service." "I don't think that's how that works." "Are you in or not?" Bill stared at the map for a long moment, then pointed at the drawing near the garage. "Is that supposed to be the face painting station or a second horse?" Molly snatched the map back. "You're driving the van."
created_at
"2026-02-08T15:27:35.972703+00:00"
Sunday, February 8th, 2026 at 3:27:35 PM GMT+00:00
model
"anthropic/claude-4.6-opus-20260205"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1578
moderation_latency
202
generation_time
18814
tokens_prompt
61
tokens_completion
639
native_tokens_prompt
75
native_tokens_completion
723
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
0.01845
router
(null)
provider_responses
0
id
"efce2503-139d-4901-83fb-d600f94f6804"
status
200
is_byok
false
latency
1576
endpoint_id
"a8708fb9-08ea-43bb-a3b1-165c87ef09df"
provider_name
"Amazon Bedrock"
model_permaslug
"anthropic/claude-4.6-opus-20260205"
api_type
"completions"
id
"gen-1770564436-EbANIKfchHX1h9pc3rBD"
upstream_id
"efce2503-139d-4901-83fb-d600f94f6804"
total_cost
0.01845
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
Evaluation details
Result
Evaluator
Details
Meta Data
35.0022%
Matches word count
n/a
neededClean
false
words
482
0.0006%
Dialogue to Total Word Ratio
Ratio: 85.48%, Deviation: 15.48%
neededClean
false
wordsTotal
482
wordsDialogue
412
17.5014%