NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-06-01KK1HH3HRTP7GMKKXYGP2ZVH9
Run Details
Model:
Inception Mercury
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
9.5s
Cost:
$0.000495
Provider:
Inception
Evaluation summary
Score
Evaluator
Details Link
35.00%
Matches word count
(details)
96.13%
Dialogue to Total Word Ratio
(details)
65.5680%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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
Moonlight slipped through the cracked blinds of the old townhouse, casting a silver halo on the dusty floor. The house was oddly quiet, save for the faint hum of a distant television. Inside, a child's party was in full swing—teddy bears perched on chairs, balloons bobbing in the air, and a crooked cake that seemed to have its own agenda. Molly and Bill stood at the back, their eyes scanning the room with the precision of seasoned thieves. They had spent the last week mapping the layout, noting the placement of the gift table and the security system that relied on a bored babysitter. The plan was simple: slip in, grab the presents, and vanish before anyone noticed. But the real challenge was to keep their composure while the giggling children and the oblivious parents made the scene feel like a circus. Molly: "Okay, Bill, we got the blueprint. The gift table is right behind the cake. We just need to slip in through the back door, grab the presents, and exit before the kids notice." Bill: "Right, but the babysitter is on a nap. We should use the smoke detector to trigger a false alarm. That will give us a window." Molly: "I already set up a tiny firecracker in the hallway. When it goes off, the alarm will go off, and the babysitter will run to the kitchen. We'll be in and out in less than a minute." Bill: "Sounds good, but we need a distraction. Maybe we can rig the balloon arch to pop. The kids will be busy trying to catch the confetti." Molly: "And the cake? We can use the frosting as a slip device. Just a quick swipe and the presents are gone." Bill: "I love your optimism. But the parents are watching the TV. We need to make sure the TV doesn't turn on during our break." Molly: "I'll plug the remote into a deadbolt. The TV will stay off. Also, I'll hide the key under the rug." Bill: "Great plan. Let's rehearse. Molly, you go first. You slip in through the back door, grab the presents, and exit. I'll wait in the hallway." Molly: "Got it. I'll be in and out before the alarm triggers." Bill: "And if the alarm goes off, we just run to the kitchen and hide behind the curtain." Molly: "Exactly. We'll be like ninjas in a circus." Bill: "And the kids will think it's a magic show. The presents will be ours, and the party will be... well, a mess." Molly: "Perfect. Let's do this." Bill: "If the kids start singing 'Happy Birthday' to the cake, we can use the chorus to our advantage. Just slip in the frosting and pretend the cake is a snowman." Molly: "And if the babysitter returns early, we can blame the cake for being too cold. She'll think it's a prank and leave the presents where they are." Bill: "And if the kids get curious, we can pretend the presents are magic beans that grow into candy." Molly: "We'll leave a trail of glitter to distract the kids."
created_at
"2026-03-06T12:21:32.351Z"
3/6/2026, 12:21:32 PM
model
"inception/mercury"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
8977
moderation_latency
(null)
generation_time
9402
tokens_prompt
77
tokens_completion
730
native_tokens_prompt
73
native_tokens_completion
655
native_tokens_completion_images
(null)
native_tokens_reasoning
0
native_tokens_cached
65
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
0.000494875
router
(null)
provider_responses
0
endpoint_id
"da1082ea-721e-460e-b330-034ae7aefc7f"
id
"chatcmpl-ebc2a379-ac35-4370-9048-0a8c73e694ae"
is_byok
false
latency
43
model_permaslug
"inception/mercury"
provider_name
"Inception"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1772799692-7GUHxleCfHHBPNXe0byy"
upstream_id
"chatcmpl-ebc2a379-ac35-4370-9048-0a8c73e694ae"
total_cost
0.000494875
cache_discount
0.000014625
upstream_inference_cost
0
provider_name
"Inception"
Evaluation details
Result
Evaluator
Details
Meta Data
35.00%
Matches word count
n/a
neededClean
false
words
518
96.13%
Dialogue to Total Word Ratio
Ratio: 69.11%, Deviation: 0.89%
neededClean
false
wordsTotal
518
wordsDialogue
358
65.5680%