NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37PRY4F63GVW5JS3ATRH7BF
Run Details
Model:
Aion 3.5 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
27.1s
Cost:
$0.005827
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
22.48%
Dialogue to Total Word Ratio
(details)
11.2421%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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 crouched behind the hydrangea bushes of 14 Maple Drive, binoculars pressed to her face, studying the enemy encampment with the focus of a general surveying a battlefield. The enemy, in this case, was a bounce house, a folding table bearing a cake shaped like a dinosaur, and approximately eleven children under the age of eight, hopped up on juice boxes and zero moral scruples. Bill lay flat in the grass beside her, wearing a Hawaiian shirt that could be seen from orbit. "You look conspicuous," Molly whispered. "I look like a dad at a barbecue. Dads are invisible at these things." "You look like a dad who lost a bet." The presents were stacked on a picnic table near the garage, a glittering mountain of wrapping paper and ribbon, completely unguarded. It was almost insulting. Molly had pulled off museum jobs with more security than this. The Louvre had lasers. This had a golden retriever named Biscuit, currently asleep in a sunbeam. "Okay," she said, unrolling a napkin with a diagram drawn in ketchup. "Phase one: you enter through the side gate dressed as the entertainment. I've booked us under the name 'Professor Twinkle's Magic Hour.' Phase two: while you distract them with the balloon animals—" "I don't know how to make balloon animals." "Bill." "I can inflate them. That's the extent of my skill set." "Then you'll inflate them with such charisma they won't notice the balloon is just a balloon." She jabbed a ketchup-drawn arrow toward the table. "Phase three: I load the presents into the wheelbarrow while everyone watches you." Bill squinted at the napkin. "Why is the wheelbarrow drawn as a tank?" "Artistic license." A shriek erupted from the party. They both ducked. A small girl in a tiara had seized another child's party bag, and negotiations had broken down into screaming and what Molly could only describe as tactical biting. Two mothers intervened. It was brutal. It was efficient. Molly felt something she hadn't felt since her days in the SAS of party crashing: respect. "These aren't normal civilians," she murmured. "These are seven-year-olds. They have no fear, no mercy, and no object permanence. You get spotted, they'll swarm you like ants on a dropped ice cream cone." "Then we go loud," Bill said, sitting up with grim determination, a sprig of hydrangea in his hair. "I walk in, I announce the magic show, and I am so mesmerizing that no one looks at the present table for twenty minutes." "You've never mesmerized anyone in your life." "I mesmerized a pigeon once." "You fell asleep on a bench and a pigeon watched you. That's not mesmerism, that's a bird making a judgment call." Molly checked her watch. The party hit its sugar-collision peak at 2:15, when the cake came out and every parental eye would be fixed on the ceremonial lighting of the candles. That was their window. Ninety seconds of darkness and frosting. She'd stolen diamonds out of vaults with worse odds. "Positions," she whispered, smearing the ketchup map into the grass. Bill stood, adjusted his shirt, and strutted toward the side gate with the confidence of a man who believed his own hype. He made it four steps before the golden retriever woke up, looked directly at him, and began to bark with the enthusiasm of a dog who had finally found his purpose in life. Every head at the party turned. "Professor Twinkle!" Bill announced, arms wide, voice cracking. "Has arrived!" A small boy in the tiara pointed at him and, in a voice of pure betrayal, screamed, "That's not a real wizard!" Molly closed her eyes. Wheelbarrow, she thought. Forget the wheelbarrow. She began calculating the fastest route back to the bushes.
created_at
"2026-09-23T17:59:47.616Z"
9/23/2026, 5:59:47 PM
model
"aion-labs/aion-3.5-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
989
moderation_latency
(null)
generation_time
27058
tokens_prompt
75
tokens_completion
1085
native_tokens_prompt
65
native_tokens_completion
963
native_tokens_completion_images
(null)
native_tokens_reasoning
152
native_tokens_cached
65
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
0.00582675
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790186387-fZ4ptqf9YTG3IYuy0RUJ"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790186387-bYYLUSeNuAbwn4P1tR7B"
upstream_id
"580cdbec7a5044b08cac6db05dc55563"
provider_responses
0
endpoint_id
"09b1fc56-3888-42bd-b7a0-926851c7f9d6"
id
"580cdbec7a5044b08cac6db05dc55563"
is_byok
false
latency
988
model_permaslug
"aion-labs/aion-3.5-20260923"
provider_name
"AionLabs"
status
200
total_cost
0.00582675
cache_discount
0.00014625
upstream_inference_cost
0
provider_name
"AionLabs"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
622
22.48%
Dialogue to Total Word Ratio
Ratio: 35.46%, Deviation: 5.46%
neededClean
false
wordsTotal
626
wordsDialogue
222
11.2421%