NC Bench
Overview
About
Tests
Models
Vendors
Compare
Run: 2026-02-24-01KJ8VRZA4GJR0JHHP29QNDCQ4
Run Details
Model:
Claude Opus 4.6 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
1m 10s
Cost:
$0.150135
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
99.9996%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 hood of their rusted minivan, squinting at what appeared to be a crayon rendering of a bouncy castle. Bill leaned over her shoulder, breathing heavily through his mouth in a way that suggested both chronic sinusitis and deep concentration. The target was simple enough: a sixth birthday party for one Emma Fitzgerald, whose parents had more money than architectural taste. Intelligence gathered from the neighborhood Facebook group suggested a pile of presents worth several hundred dollars, minimum. There would be a magician at three o'clock, which meant every child would be distracted watching a man in a cape butcher card tricks. "We go in during the magic show," Molly whispered. Bill nodded, then frowned. "What about the dog?" The Fitzgeralds owned a golden retriever named Captain Biscuit who, according to surveillance conducted from behind a mailbox, loved absolutely everyone. This was both a blessing and a problem, since the animal would almost certainly follow them around demanding belly rubs and drawing attention. "Bring some treats," Molly said. "The good ones. Bacon flavor." Bill scribbled this on his palm with a dying ballpoint pen and smiled like a man who'd finally found his calling.
created_at
"2026-02-24T22:19:35.115Z"
2/24/2026, 10:19:35 PM
model
"anthropic/claude-4.6-opus-20260205"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2161
moderation_latency
220
generation_time
69767
tokens_prompt
77
tokens_completion
788
native_tokens_prompt
92
native_tokens_completion
5987
native_tokens_completion_images
(null)
native_tokens_reasoning
462
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
0.150135
router
(null)
provider_responses
0
endpoint_id
"a8708fb9-08ea-43bb-a3b1-165c87ef09df"
id
"6464cd56-eebf-413f-a94d-209a83cd579c"
is_byok
false
latency
2159
model_permaslug
"anthropic/claude-4.6-opus-20260205"
provider_name
"Amazon Bedrock"
status
200
api_type
"completions"
id
"gen-1771971575-SYtkRQxUlueuynTNcklV"
upstream_id
"6464cd56-eebf-413f-a94d-209a83cd579c"
total_cost
0.150135
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
200
100.00%
Dialogue to Total Word Ratio
Ratio: 9.45%, Deviation: 0.55%
neededClean
false
wordsTotal
201
wordsDialogue
19
99.9996%