NC Bench
Overview
About
Tests
Models
Model
Claude Opus 4.6
Test
Dialogue tags
Scenario
Write 200 words with 90% dialogue
Run ID
2026-02-08-01KGYXGMP5YXFTXM22BD02XP6J
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
Molly spread the crayon-drawn map across the dashboard and tapped the big pink house. "The party starts at two. That gives us a forty-minute window while they're doing the piñata in the backyard." "I don't like piñatas," Bill said. "One hit me in the face once." "Focus, Bill." "I'm just saying. Papier-mâché is deceptively heavy." "The presents will be stacked on a folding table in the living room. We go in through the side door, load the bags, and we're out before they sing Happy Birthday." "What if a kid sees us?" "We're wearing the clown costumes." "I'm not wearing the clown costume, Molly." "You're wearing the clown costume." "Last time I wore that thing, a toddler bit me." "You'll survive. Now, the mother ordered a bouncy castle, which means every adult will be distracted watching kids bounce into each other. It's the perfect cover." "What kind of presents are we talking here?" "LEGOs, Bill. Mountains of LEGOs." "LEGOs hurt when you step on them." "Everything hurts you, Bill. Are you in or not?" Bill sighed and picked up the red foam nose. "I'm in."
created_at
"2026-02-08T15:22:02.790625+00:00"
Sunday, February 8th, 2026 at 3:22:02 PM GMT+00:00
model
"anthropic/claude-4.6-opus-20260205"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1261
moderation_latency
253
generation_time
8670
tokens_prompt
61
tokens_completion
254
native_tokens_prompt
75
native_tokens_completion
309
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
0.0081
router
(null)
provider_responses
0
id
"ba275558-d95d-473d-8d1c-cbb4651f137b"
status
200
is_byok
false
latency
1261
endpoint_id
"a8708fb9-08ea-43bb-a3b1-165c87ef09df"
provider_name
"Amazon Bedrock"
model_permaslug
"anthropic/claude-4.6-opus-20260205"
api_type
"completions"
id
"gen-1770564113-tyAGiBlaLafrphsiwDNB"
upstream_id
"ba275558-d95d-473d-8d1c-cbb4651f137b"
total_cost
0.0081
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
Evaluation details
Result
Evaluator
Details
Meta Data
81.2727%
Matches word count
n/a
neededClean
false
words
188
97.8717%
Dialogue to Total Word Ratio
Ratio: 86.17%, Deviation: 3.83%
neededClean
false
wordsTotal
188
wordsDialogue
162
89.5722%