NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-18-01KHQ922A8VRKC33PSHX2ZH1TY
Run Details
Model:
Claude Sonnet 4.6
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
9.1s
Cost:
$0.005430
Provider:
Anthropic
Evaluation summary
Score
Evaluator
Details Link
51.93%
Matches word count
(details)
93.42%
Dialogue to Total Word Ratio
(details)
72.6704%
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
"I still can't believe we're doing this," Molly whispered, peering over the garden fence. "It's a victimless crime," Bill said. "They're seven year olds, Bill." "Who have too many presents. We're basically teaching them about loss. It's character building." "Is that what we're calling it now?" "We're philanthropists." "We're wearing black turtlenecks outside a bouncy castle." "Camouflage." "It's a pink bouncy castle." "Then we're slightly visible philanthropists. Now, the presents are by the clown." "There's a clown?" "Don't worry about the clown." "Bill, I'm absolutely worrying about the clown." "He's between us and forty wrapped gifts. Stay focused." "What's in the bags anyway?" "Lego, mostly. Some dolls. A remote control car that I actually want for myself." "This is rock bottom." "Rock bottom was the casino job." "Fair point." "On three we go over the fence, grab the presents, ignore the clown, and run." "And if someone spots us?" "Tell them we're the entertainment." "We're not the entertainment." "Then run faster. Ready?" "Absolutely not." "One, two—" "Wait, is that clown looking at us?" "He's always looking at us, Molly. That's what clowns do. Go."
created_at
"2026-02-18T02:25:35.081772+00:00"
2/18/2026, 2:25:35 AM
model
"anthropic/claude-4.6-sonnet-20260217"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
789
moderation_latency
(null)
generation_time
9042
tokens_prompt
77
tokens_completion
341
native_tokens_prompt
75
native_tokens_completion
347
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
0.00543
router
(null)
provider_responses
0
id
"msg_01RMRYXXAqzgsfuYkYRy5mXp"
status
200
is_byok
false
latency
789
endpoint_id
"6d408764-7dd7-4626-bb87-a6cc1589bc86"
provider_name
"Anthropic"
model_permaslug
"anthropic/claude-4.6-sonnet-20260217"
api_type
"completions"
id
"gen-1771381524-UYAljdg7DUIQ61QgGiQA"
upstream_id
"msg_01RMRYXXAqzgsfuYkYRy5mXp"
total_cost
0.00543
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Anthropic"
Evaluation details
Result
Evaluator
Details
Meta Data
51.93%
Matches word count
n/a
neededClean
false
words
184
93.42%
Dialogue to Total Word Ratio
Ratio: 95.11%, Deviation: 5.11%
neededClean
false
wordsTotal
184
wordsDialogue
175
72.6704%