NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-07-01KWZ37K3E3JR745TEVXJE6HF0
Run Details
Model:
Aion 3.0 Mini
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
23.2s
Cost:
$0.001177
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
0.21%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.1070%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% 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 adjusted his monocle, which he’d bought for exactly this job. “The target is the ‘Magical Mermaid Princess Party Spectacular.’ The swag is the gift table.” Molly, dressed in a garish rainbow clown wig, was already sweating. “I still don’t see why we have to do this. The risk-reward ratio is—“ “Off the charts!” Bill hissed, pulling a blueprint from his jacket. It was a crayon drawing of a bouncy castle, with an ‘X’ marked on the snack table. “We’re talking brand new, unboxed toys, Molly. Think of the resale value on an unassembled Lego pirate ship.” “I’m thinking of the traumatic screech of a six-year-old who just lost their FurReal puppy.” “That’s called ‘market disruption.’ Now, focus. I’ve cased the joint. The party starts at two. The mother, codename ‘Samantha,’ is a sucker for balloon artists. We enter at 2:17, right when the sword-swallower is doing the big finale.” Molly pulled the clown wig off her head. “Bill, the ‘sword-swallower’ is a man wearing a cardboard sword and a party hat. The finale is him eating a piece of string cheese in one bite.” “A decoy! Brilliant. They’ll be watching him, not me.” Bill pointed a crumb-covered finger at the blueprint. “I’ll enter via the rear hedge, dressed as a mime. My silent, pained expressions will lull the children into a false sense of security while I grab the gift bags.” “And what will I be doing while you’re doing your Marcel Marceau impression?” Bill stroked his chin, a glob of jelly from his doughnut landing on the blueprint. “You will be my distraction. A rival clown. You will juggle three beanbags poorly. The children will laugh at you, pointing. When they see a mime silently stealing their presents, they’ll assume it’s part of your comedy routine.” “So you want me to be a clown, juggle badly, and take the heat for a mime who’s actually a thief.” “Precisely! It’s the perfect crime.” “And what if I say no? What if I just walk away from this inflatable-furnished, juice-box-fueled den of chaos?” “You can’t. I already told them you were coming.” Bill smiled, revealing a piece of the doughnut stuck to his tooth. “I told Samantha, the mother, that you’re ‘Sparkles the Cake-pooping Unicorn.’ The kids are expecting you. If you don’t show, they’ll be disappointed. You hate disappointing children.” Molly stared at him, her face a mask of pure, undiluted contempt. “I hate you.” “Hate is a strong word for a partner. Now, about the balloon animal part of your act. I need you to make a giraffe. Because when the cops show up, I want them to be confused.” “Confused about what?” “The giraffe will be a red herring. They’ll chase the person holding the giraffe. That’s you.” Molly put the clown wig back on. It felt heavier. “I’m going to need a better cut of the pirate ship.” “Deal. Now, are we ready to commit the perfect heist?” “We’re going to go steal a bunch of presents from toddlers. Perfect is not the word I’d use.” “It’s a start.” Bill said. “And a good start, is half the battle. Let’s get our clown on.”
created_at
"2026-07-07T20:10:09.918Z"
7/7/2026, 8:10:09 PM
model
"aion-labs/aion-3.0-mini-20260707"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
928
moderation_latency
(null)
generation_time
23087
tokens_prompt
77
tokens_completion
863
native_tokens_prompt
66
native_tokens_completion
808
native_tokens_completion_images
(null)
native_tokens_reasoning
11
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
"stop"
service_tier
(null)
usage
0.0011774
router
(null)
provider_responses
0
endpoint_id
"0ad7ab68-a351-4a96-b2ef-fa8fe13adae3"
id
"360ea2447afa463da6590040d73ddfaa"
is_byok
false
latency
927
model_permaslug
"aion-labs/aion-3.0-mini-20260707"
provider_name
"AionLabs"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1783455009-TA1sfR8IPpC7eO5B3vBv"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783455009-d1jqH6N3tCiAR4ypB8zB"
upstream_id
"360ea2447afa463da6590040d73ddfaa"
total_cost
0.0011774
cache_discount
(null)
upstream_inference_cost
0
provider_name
"AionLabs"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
0.21%
Matches word count
n/a
neededClean
false
words
528
0.00%
Dialogue to Total Word Ratio
Ratio: 79.70%, Deviation: 29.70%
neededClean
false
wordsTotal
537
wordsDialogue
428
0.1070%