NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-18-01KHQ9180HWHPJTX1RVCS0FHNY
Run Details
Model:
Claude Sonnet 4.6
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
8.7s
Cost:
$0.004230
Provider:
Anthropic
Evaluation summary
Score
Evaluator
Details Link
99.38%
Matches word count
(details)
52.27%
Dialogue to Total Word Ratio
(details)
75.8259%
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
The storage room behind Funtime Frankie's Party Palace smelled like discount cake frosting and broken dreams. Molly spread the hand-drawn floor plan across a cardboard box, weighing down the corners with a juice box and somebody's lost retainer she absolutely should not have touched. Bill stood beside her, wearing a mustache disguise that fooled absolutely nobody, least of all himself. The birthday party was in full swing next door. Seventeen children screaming at a magician who had clearly made better life choices than either of them. The presents table sat near the emergency exit, groaning under the weight of approximately forty boxes wrapped in dinosaur and princess paper. Molly had counted them twice during reconnaissance, disguised as a balloon delivery woman. Bill shifted nervously from foot to foot, his mustache sliding slowly toward his chin like a sleeping caterpillar. "Ready?" Molly whispered. "No," Bill said. She appreciated his honesty. Neither was she. The plan had seventeen steps, and step four involved befriending the party clown, which she considered the most dangerous component of any operation she had ever attempted in twelve years of moderately successful criminal activity. She rolled up the floor plan anyway.
created_at
"2026-02-18T02:25:07.326263+00:00"
2/18/2026, 2:25:07 AM
model
"anthropic/claude-4.6-sonnet-20260217"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
402
moderation_latency
(null)
generation_time
8574
tokens_prompt
77
tokens_completion
321
native_tokens_prompt
75
native_tokens_completion
267
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
0.00423
router
(null)
provider_responses
0
id
"msg_018ZMyE72P12rFPG1mbtyiwE"
status
200
is_byok
false
latency
402
endpoint_id
"6d408764-7dd7-4626-bb87-a6cc1589bc86"
provider_name
"Anthropic"
model_permaslug
"anthropic/claude-4.6-sonnet-20260217"
api_type
"completions"
id
"gen-1771381497-xYYx6lYd7nj4oxcIL2Tn"
upstream_id
"msg_018ZMyE72P12rFPG1mbtyiwE"
total_cost
0.00423
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Anthropic"
Evaluation details
Result
Evaluator
Details
Meta Data
99.38%
Matches word count
n/a
neededClean
false
words
195
52.27%
Dialogue to Total Word Ratio
Ratio: 1.03%, Deviation: 8.97%
neededClean
false
wordsTotal
195
wordsDialogue
2
75.8259%