NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-01-01KWE9DH94Y5M0P1MAW17PPCQP
Run Details
Model:
Claude Sonnet 5 (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
9.3s
Cost:
$0.004836
Provider:
Amazon Bedrock
Evaluation summary
Score
Evaluator
Details Link
20.19%
Matches word count
(details)
36.79%
Dialogue to Total Word Ratio
(details)
28.4888%
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
"Bill, tell me you've got the plan." "I've got the plan." "Tell me the plan." "We go in through the bounce house." "The bounce house." "It's inflatable, Molly. It's got structural weaknesses." "It's a bounce house, Bill, not the Berlin Wall." "Every fortress has a weak point!" "It's for six-year-olds!" "Which is why nobody will suspect us." "Nobody will suspect us because we're forty-year-old adults crawling into a bounce house?" "Exactly. Genius, right?" "I hate this. I hate this so much." "You said that about the bank job." "The bank job worked!" "This'll work too! We wait till they do the piñata, everyone's distracted, we slide in, grab the presents, slide out." "Slide out of a bounce house. With an armful of presents." "You're overthinking it." "There's a clown, Bill. Did you account for the clown?" "Clowns are cowards. Basic thief knowledge." "That's not a real rule." "It's rule forty-two." "There is no rulebook!" "There's absolutely a rulebook." "Fine. Fine! Bounce house. Piñata. Clown's a coward. Anything else?" "Don't eat the cake. It's got a tracking chip." "WHAT?" "Kidding. Probably."
created_at
"2026-07-01T07:31:10.765Z"
7/1/2026, 7:31:10 AM
model
"anthropic/claude-sonnet-5-20260630"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2000
moderation_latency
(null)
generation_time
9154
tokens_prompt
77
tokens_completion
327
native_tokens_prompt
93
native_tokens_completion
465
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)
num_fetches
(null)
web_search_engine
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"end_turn"
service_tier
"default"
usage
0.004836
router
(null)
provider_responses
0
endpoint_id
"37b16f45-eeae-45f3-9b85-2137f7c6d61b"
id
"msg_01QF9PTJMsQHEtxs4Ev82kri"
is_byok
false
latency
2000
model_permaslug
"anthropic/claude-sonnet-5-20260630"
provider_name
"Amazon Bedrock"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.17.0; linux; x64))"
http_referer
(null)
request_id
"req-1782891070-TH1NuICUzOYy74Bo9jPE"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1782891070-Bd93c6twxvGGN8snKb7V"
upstream_id
"msg_01QF9PTJMsQHEtxs4Ev82kri"
total_cost
0.004836
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
20.19%
Matches word count
n/a
neededClean
false
words
180
36.79%
Dialogue to Total Word Ratio
Ratio: 100.00%, Deviation: 10.00%
neededClean
false
wordsTotal
185
wordsDialogue
185
28.4888%