NC Bench
Overview
About
Tests
Models
Model
Claude Sonnet 4.5
Test
Dialogue tags
Scenario
Write 200 words with 10% dialogue
Run ID
2026-02-08-01KGYX0K87R7VDYS8Y9DAR1XCS
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 examined the birthday banner through her binoculars while Bill nervously tapped his fingers against the steering wheel. The suburban house burst with balloons and streamers, a fortress of innocent celebration they were about to violate for profit. Bill cleared his throat and shifted uncomfortably in his seat. His conscience had been bothering him since they'd accepted this ridiculous job. Molly lowered the binoculars and studied the party's chaotic rhythm. Children zigzagged across the lawn wielding foam swords while exhausted parents clutched coffee cups like lifelines. The present table sat magnificently unguarded near the patio, wrapped boxes glittering in the afternoon sun. She pulled out her phone and scrolled through their client's specifications. Apparently, someone's ex-spouse had extremely expensive taste in children's gifts, and revenge required intercepting them before little Timmy could tear into that limited edition whatever-it-was. Bill drummed his fingers faster. "This feels wrong." "Two grand says otherwise." "Still wrong." Molly pocketed her phone and checked her watch. The clown was scheduled to arrive in ten minutes, which would provide the perfect distraction. She reached for the door handle, then hesitated. Bill was right, but rent was righter.
created_at
"2026-02-08T15:13:15.989806+00:00"
Sunday, February 8th, 2026 at 3:13:15 PM GMT+00:00
model
"anthropic/claude-4.5-sonnet-20250929"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1983
moderation_latency
151
generation_time
7951
tokens_prompt
61
tokens_completion
252
native_tokens_prompt
75
native_tokens_completion
288
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
0.004545
router
(null)
provider_responses
0
id
"116036ba-fe25-4d0f-afb1-7b4cfab51746"
status
200
is_byok
false
latency
1983
endpoint_id
"d8bee848-f020-454a-b2be-4a7ce868f99f"
provider_name
"Amazon Bedrock"
model_permaslug
"anthropic/claude-4.5-sonnet-20250929"
api_type
"completions"
id
"gen-1770563587-FBHOV3H1ssuyE8xV3tAn"
upstream_id
"116036ba-fe25-4d0f-afb1-7b4cfab51746"
total_cost
0.004545
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Amazon Bedrock"
Evaluation details
Result
Evaluator
Details
Meta Data
95.9868%
Matches word count
n/a
neededClean
false
words
192
92.3438%
Dialogue to Total Word Ratio
Ratio: 4.69%, Deviation: 5.31%
neededClean
false
wordsTotal
192
wordsDialogue
9
94.1653%