NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37PBZX17YCR1WH2HGYBNR3N
Run Details
Model:
Aion 3.5 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
20.4s
Cost:
$0.005593
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0004%
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
Molly spread the blueprint across the kitchen table, weighting the corners with a bag of gummy worms and a novelty mug that read World's Okayest Criminal. "Alright," she said. "The party starts at two. We go in at two fifteen, when the cake comes out. Every parent in that yard will be staring at a five-year-old trying to blow out a candle with the lung capacity of a hamster." "That's our window," Bill said, nodding sagely. "That's our window. Now, the presents are on a folding table by the garage. We grab the big ones first. The big ones are always the good ones. Nobody wraps a sock in a box the size of a dishwasher." "Unless it's a trampoline," Bill said. "It's never a trampoline, Bill." "It could be a trampoline." "It's a sock. It's always a sock, and the box is full of packing peanuts, and somewhere a parent is laughing at us." Bill chewed his thumb. "What about the clown?" "What about the clown?" "There's a clown. I saw the flyer at the community center. 'Bubbles the Clown, two to four.'" "Fine. We handle the clown." "How do you handle a clown?" "Same way you handle anyone. You don't make eye contact and you walk with purpose." Bill shook his head. "Clowns can smell fear, Molly. Everyone knows that." "Then don't be afraid. Be a birthday clown yourself. That's the disguise, you idiot. We go in as entertainers. You're a clown, I'm a balloon artist." "I don't know how to make balloons." "You make a snake. Every balloon animal you're asked for, you say, 'Funny story, it's a snake.' Kids love snakes. Kids don't know what a swan is anyway." "Okay, okay." Bill leaned over the blueprint. "What's this rectangle here?" "The bouncy castle." "We're stealing a bouncy castle?" "We are not stealing a bouncy castle. You said that last time and we spent four hours in a moonwalk with no exit strategy." "It was a good castle," Bill said quietly. "It was evidence, Bill. It had a police sergeant's daughter in it." Molly tapped the table. "Focus. We're in and out in eleven minutes. You take the left side of the present table, I take the right. We meet at the hedge." "What about the piñata?" "What about it?" "Is there treasure in the piñata?" "There's candy in the piñata." "Candy is treasure." "Candy is a felony multiplier, Bill. We're not adding candy to a grand theft presents charge." Bill sighed and looked at the ceiling. "I just think, if we're doing the job anyway, the piñata's right there." "The piñata is a distraction. The piñata is a room full of children holding bats. You know what happens when a piñata breaks? Chaos. Screaming. Blindfolded children swinging at anything that moves, including a middle-aged clown with a pillow case full of wrapped toys." "So we skip the piñata." "We skip the piñata." "What if the piñata has an alarm?" "Why would a piñata have an alarm?" "What if it's not a real piñata? What if it's a decoy piñata, and the real presents are inside it?" Molly stared at him for a long moment. "You know what? Fine. Fine. We take the piñata. We take the whole piñata, and if there's a sock in it, that's on you." Bill grinned and pulled on his rainbow wig. "Two fifteen," she said. "Cake time." "Snake time," he corrected, and honked his nose.
created_at
"2026-09-23T17:52:43.442Z"
9/23/2026, 5:52:43 PM
model
"aion-labs/aion-3.5-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
769
moderation_latency
(null)
generation_time
20305
tokens_prompt
75
tokens_completion
982
native_tokens_prompt
65
native_tokens_completion
924
native_tokens_completion_images
(null)
native_tokens_reasoning
148
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
0.00559275
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790185963-XlvwOHzB6tQznDhr9z4V"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790185963-Bzh8yu2DNDQakDwTQVpU"
upstream_id
"fa00249e76e8438a8d5a6b6f71c6f24e"
provider_responses
0
endpoint_id
"09b1fc56-3888-42bd-b7a0-926851c7f9d6"
id
"fa00249e76e8438a8d5a6b6f71c6f24e"
is_byok
false
latency
769
model_permaslug
"aion-labs/aion-3.5-20260923"
provider_name
"AionLabs"
status
200
total_cost
0.00559275
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
574
0.00%
Dialogue to Total Word Ratio
Ratio: 85.27%, Deviation: 15.27%
neededClean
false
wordsTotal
577
wordsDialogue
492
0.0004%