NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-24-01M39XGME03NM4JTYYEY4SDXYK
Run Details
Model:
Aion 3.5 Mini (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
29.0s
Cost:
$0.002243
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
99.92%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
49.9595%
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
The van sat across the street from 42 Maple Drive, where a banner reading HAPPY 7TH BIRTHDAY TIMMY flapped in the breeze like it was mocking them personally. Molly adjusted her binoculars. "Twelve kids. Fourteen adults. One bouncy castle." "The bouncy castle is a problem," Bill said. "It's got walls. Walls made of air." "It's a bouncy castle, Bill, not Alcatraz." "Alcatraz didn't try to launch me into a tree at my nephew's birthday." "That was one time, and it was your fault for eating the entire cake first." Bill squinted through the windshield. "What's the plan? We walk in wearing what, exactly? We're not the mommy-and-me demographic." Molly reached into the back and produced a clown costume. Then a second one. "No," Bill said. "You already have a name picked out. Jangles." "I said no to that name specifically." "Jangles the Clown," Molly continued, "arrives at two-fifteen, juggles badly, and performs a magic trick where the presents disappear." "The presents don't actually disappear, Molly. They disappear into our van." "That's the twist. Kids love a twist." "Kids love presents. Kids do not love their presents being stolen by a sad man named Jangles." Molly checked her watch. "Cake's at two. The kids will be sugar-shocked, staring at the birthday boy like he's royalty. Nobody watches the gift table during cake. Oldest rule in heisting." "Is it? Is that a rule?" "It's my rule. I just made it, but it's true." Bill drummed his fingers on the wheel. A woman in a party hat was taping a laminated sign to the front door. "She's laminating party signage," Molly said. "This family is organized. We have to be better." "Organized people label things. Organized people notice missing presents." "Which is why we're not taking all of them." Bill turned slowly. "Since when do we have ethics?" "It's not ethics, it's math. We take eight of twelve. Four remain, parents assume the kids opened them already, kids assume the parents hid them for later. Everybody's confused, nobody's certain, and confusion is our getaway car." Bill was quiet for a long moment. "That's actually evil," he said. "Evil and clean." "I know." "Can I still be Jangles?" "You want to be Jangles?" "I've grown attached. It's a good name. It has bounce." Molly pulled the red foam nose from its packaging and rolled it between her fingers. "Fine. Jangles it is. But if you get launched into a tree again, I'm leaving you there." "Deal. One condition." "What?" "We leave the pony. I saw a pony in the backyard. Nobody steals a pony, Molly. That's a line, and it's the only line I have left." "The pony wasn't even on the list." "It wasn't on the list because you knew I'd say no." "I knew you'd say yes, which is worse," Molly said, sliding her nose on. "Okay. Cake in ninety seconds. Get your balls, Jangles." Bill grabbed his juggling set and took a deep breath. "For Timmy," he whispered. "We're robbing Timmy." "For nobody, then."
created_at
"2026-09-24T14:36:04.431Z"
9/24/2026, 2:36:04 PM
model
"aion-labs/aion-3.5-mini-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1429
moderation_latency
(null)
generation_time
28980
tokens_prompt
75
tokens_completion
1721
native_tokens_prompt
65
native_tokens_completion
1594
native_tokens_completion_images
(null)
native_tokens_reasoning
959
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
0.0022433
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790260564-fCupsbo78ZjnjT0FzIVe"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790260564-KKreOm7VMCGRCAFgAX9p"
upstream_id
"536a75ed39f149fbb18aaa9c1a33f727"
provider_responses
0
endpoint_id
"60ac3bdc-90ec-4769-9193-2c00b4d79562"
id
"536a75ed39f149fbb18aaa9c1a33f727"
is_byok
false
latency
1429
model_permaslug
"aion-labs/aion-3.5-mini-20260923"
provider_name
"AionLabs"
status
200
total_cost
0.0022433
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
99.92%
Matches word count
n/a
neededClean
false
words
503
0.00%
Dialogue to Total Word Ratio
Ratio: 74.36%, Deviation: 24.36%
neededClean
false
wordsTotal
507
wordsDialogue
377
49.9595%