NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-07-01KWZ5TY22AZN3WBRDGQYEADSG
Run Details
Model:
Aion 3.0
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
18.4s
Cost:
$0.006171
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
68.10%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
34.0511%
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
The basement of Mama Lucrezia's Pizza Parlor smelled of garlic and poor decisions. Molly hunched over a table covered in blueprints, crayon drawings, and what appeared to be a suspiciously detailed map of a bounce house. Bill sat across from her, his massive frame threatening to buckle the plastic chair beneath him as he studied the plans through bifocals that had seen better decades. "Alright," Molly whispered, though they were alone. "The target is Timmy Henderson's seventh birthday party. Saturday, two o'clock sharp. I've cased the joint three times this week." "You went to a stranger's kid's party three times?" "I told them I was the balloon sculptor. Turns out I'm quite good at making poodles." Bill scratched his beard, something molten likely stuck in it from the calzone he'd consumed an hour prior. "This seems risky. Can't we just rob a bank like normal people?" Molly slapped the table with enough force to rattle their sodas. "Do you know what's at this party, Billy? A PlayStation 5. Lego sets still in their boxes. A mountain of Amazon packages that would make Jeff Bezos blush." "Those are all addressed to a seven-year-old." "That kid didn't even earn them. I have to steal for a living. Entitlement runs in families." Bill adjusted his bifocals, the lens fogging from his breath. "So what's the plan?" Molly unrolled a photograph with the flourish of a magician producing a rabbit. "I infiltrate as the entertainment, the 'Amazing Mollydini,' while you enter through the back fence disguised as a delivery driver. Once the children are distracted by my magic show, you bag the presents and make a break for the van." "What's the van look like?" "A unmarked white van I borrowed from my brother." "Your brother who drives an ice cream truck?" "Same vehicle, different signage. I scraped off the decals. Kids won't suspect a thing." Bill leaned back, processing. "Let me get this straight. I'm supposed to walk into a child's birthday party, steal their presents while you distract them with magic tricks, and escape in a former ice cream truck?" "When you say it out loud it sounds wrong." "It sounds wrong because it is wrong, Molly." Molly sighed. "Fine. I'll sweeten the deal. Seventy-thirty split." "Sixty-forty." "Done. But you're carrying the heavy stuff." Bill's silence was his acceptance. "One more thing," Molly added. "There will be a piñata. I need you to crack that thing open during my grand finale. The confusion will provide the perfect cover for your exit." "You want me to assault a piñata?" "I want you to decimate it. Show no mercy." "Against candy-filled paper mâché." "Exactly. Can you handle it?" Bill cracked his knuckles, a sound like snapping lumber. "For sixty percent of a PlayStation, I'll destroy that piñata and everything it stands for." Molly grinned. "That's the spirit. Now, let's practice your delivery driver impression. You sound too much like a criminal." "I am a criminal." "Exactly. Sound less like yourself." Bill cleared his throat, straightened his posture. "Package delivery for, uh, small child." "Better," Molly said. "Much better."
created_at
"2026-07-07T20:55:40.88Z"
7/7/2026, 8:55:40 PM
model
"aion-labs/aion-3.0-20260707"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
743
moderation_latency
(null)
generation_time
18271
tokens_prompt
77
tokens_completion
1166
native_tokens_prompt
67
native_tokens_completion
995
native_tokens_completion_images
(null)
native_tokens_reasoning
331
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
0.006171
router
(null)
provider_responses
0
endpoint_id
"8c8f827d-2bef-468a-8f9e-8e27b470de5e"
id
"18006615ccaf4e52b24f381f350f4aac"
is_byok
false
latency
743
model_permaslug
"aion-labs/aion-3.0-20260707"
provider_name
"AionLabs"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1783457740-kZqRbl5ltrBDAz0gSDcZ"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783457740-dwliLEpWsARsmjPdDk9P"
upstream_id
"18006615ccaf4e52b24f381f350f4aac"
total_cost
0.006171
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
68.10%
Matches word count
n/a
neededClean
false
words
514
0.00%
Dialogue to Total Word Ratio
Ratio: 70.13%, Deviation: 40.13%
neededClean
false
wordsTotal
519
wordsDialogue
364
34.0511%