NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-24-01M39X12PTBYA9KS4AR57910D9
Run Details
Model:
Aion 3.5 Mini (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
60.0s
Cost:
$0.004381
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
95.99%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
47.9934%
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 adjusted the fake mustache glued to her upper lip, wondering for the hundredth time how her life had spiraled from casing art museums in Vienna to sitting in a rusted van outside a suburban split-level with a man wearing googly eyes on his ski mask. Across the street, a banner reading "HAPPY BIRTHDAY CHAD!" flapped cheerfully in the breeze, mocking her. "Alright, listen up," Bill said, slapping a rolled-up piece of construction paper onto the dashboard. "I've cased the joint. We're looking at a two-story colonial, minimal security, but heavy on the frosting." Molly unrolled the paper. It was a crayon drawing of a house, complete with a sun wearing sunglasses and a dog labeled "MAX" in shaky letters. "Bill, this is a drawing a five-year-old made." "That's the intel, Molly. The five-year-old is our inside man. Name's Timmy. He owes me a favor." "He's five. He doesn't have favors. He has chickenpox." "Details. The point is, the target is the living room. We go in through the doggy door—" "The doggy door." "It's a tight squeeze, but I've been doing my squats." Molly pinched the bridge of her nose, smearing glitter that had somehow migrated from her unicorn-horn ski mask. "Let me get this straight. You want us—two of the most wanted cat burglars in the tri-state area—to break into a seven-year-old's birthday party through a doggy door to steal presents?" "Wrapped presents, Molly. Pristine. Untouched. We're talking a Lego Death Star, a bike with a basket, and at least three remote-controlled dinosaurs." "Why?" Bill leaned in close, his googly eyes wobbling with manic intensity. "Because the kid who lives there is having a better birthday than my nephew, and I'll be damned if I let that stand." "That's the dumbest motive I've ever heard." "It's called a grudge, Molly. It's very professional." She stared out the window. Balloons shaped like dinosaurs bobbed near a rented bouncy castle. Somewhere, a child was already screaming, though whether from joy or a sugar crash was anyone's guess. "The kid's name is Chad, Bill. You're robbing a Chad." "Chad is a lie cooked up by the parents. It's all a front. The bouncy castle is a fortress." "It's a bouncy castle." "It has walls, Molly." Molly sighed, pulling her mask down fully. The unicorn horn caught on the headrest. "Fine. But I'm taking the bike. I need a new way to get to the bus stop since you totaled ours." "That's the spirit. Now remember, if the clown spots us, we abort. I don't trust a guy wearing that much hairspray. He's definitely ex-military." "He's a clown, Bill." "They're always ex-military. It's the shoes." Bill checked his watch, straightened his mask, and grabbed a pillowcase emblazoned with a cartoon ghost. "Ready?" "Do I have a choice?" "No. It's Tuesday. We always work Tuesdays." Molly groaned and opened the van door. "If we get arrested, I'm telling the cops it was your idea." "If we get arrested by a guy in a mascot costume, I'll eat the pillowcase."
created_at
"2026-09-24T14:27:34.76Z"
9/24/2026, 2:27:34 PM
model
"aion-labs/aion-3.5-mini-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1435
moderation_latency
(null)
generation_time
59935
tokens_prompt
75
tokens_completion
3277
native_tokens_prompt
65
native_tokens_completion
3121
native_tokens_completion_images
(null)
native_tokens_reasoning
2513
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
0.0043811
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790260054-ENKdP6xUCWD8Qfku6imi"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790260054-x5XHDqR8Gzc4RS0KymIM"
upstream_id
"76c27db1cee64fd38cb4b0788ac6591e"
provider_responses
0
endpoint_id
"60ac3bdc-90ec-4769-9193-2c00b4d79562"
id
"76c27db1cee64fd38cb4b0788ac6591e"
is_byok
false
latency
1435
model_permaslug
"aion-labs/aion-3.5-mini-20260923"
provider_name
"AionLabs"
status
200
total_cost
0.0043811
cache_discount
0.0000338
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
95.99%
Matches word count
n/a
neededClean
false
words
508
0.00%
Dialogue to Total Word Ratio
Ratio: 62.04%, Deviation: 32.04%
neededClean
false
wordsTotal
519
wordsDialogue
322
47.9934%