NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-24-01M39W5KTAFPFD0FNCDRYKHKZ1
Run Details
Model:
Aion 3.5 Mini (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
3m 27s
Cost:
$0.014176
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
0.49%
Matches word count
(details)
7.83%
Dialogue to Total Word Ratio
(details)
4.1609%
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
The van smelled like fudge ripple and gunpowder. Molly spread the blueprints across the dashboard, pinning one corner with a coffee can of lockpicks. "Okay. Henderson kid's seventh birthday. Saturday, two p.m., suburban cul-de-sac. Forty presents, minimum." Bill leaned in and squinted. "Why are the blueprints drawn in crayon?" "Because I stole them from the kid's coloring folder. Keep up." "It's a stick figure labeled 'MY HOUSE' with arrows pointing to 'SNACKS' and 'WHERE DAD HIDES.'" Bill flipped it over. The back was a dinosaur. "That's the good stuff, Bill. Intel is intel." "Intel from a first grader." "Better access than us. Now—obstacles. One: the clown." "Clowns are a hard no, Molly." "He's not a real clown. He's the dad in a wig. I watched him practice through the fence. He dropped all three pins on his own foot." "Two: bouncy castle. Kids go in, they don't come out." "We go around it. Three: the pinata." "Ah, the pinata. Candy vault." "It's a decoy. The real goods are on the gift table under the banner." Bill paced the length of the van, chewing a toothpick into splinters. "What about the dog?" "What dog?" "Every suburban house has a dog. It's the law." "Goldendoodle. Named Sprinkles. Licks intruders." "We crawl under the slip-and-slide, cut through the hydrangeas, and distract the dad-clown with a balloon swan. He can't do swans. His portfolio is all lumpy dogs." "And while he's crying into his balloons, you lift the presents." "I lift the presents. You handle Sprinkles." "With what?" "Your famous dog whispering." Bill stopped pacing. "My famous dog whispering is me saying 'hey buddy' in a high voice." "It worked on the guard dog at the jewelry store." "He followed me home. We don't talk about Gerald." "Then it'll work on Sprinkles." Bill sat down heavily on a crate of stolen fireworks. "One flaw. Big one. Huge." "Name it." He pointed at the banner sketch, mournfully. "They're kids' presents, Molly. Seventh birthday. Once in a lifetime." "Every year there's a new seventh birthday. It's an industry." "But the banner has Tyler's name on it." "We're professionals, Bill." "Professionals with a conscience clause." Molly gave him a look that had launched a thousand botched jobs. "Since when?" "Since the bingo hall. Mrs. Kowalski slapped me with her purse." "That's not conscience, that's trauma." "Look—new plan. We don't steal the presents. We steal the cake. The three-tier ice cream monstrosity." Molly stared at him, then slowly rolled up the crayon blueprints. "You're proposing a cake heist." "A cake heist." "That's the dumbest idea that's ever worked. What's our exit?" "Same as planned. Slip-and-slide, hydrangeas, van." "And Sprinkles?" "Hey buddy." He demonstrated the voice, which was genuinely impressive. Molly grabbed the keys and started the engine. "Fine. But if a seven-year-old chases us down the street for cake, you're doing the running." "Deal. But we leave a dollar. Kids love dollars." "Bill." "Tip culture, Molly. It's a party." She pulled out of the alley, checking the rearview mirror out of pure habit. "One dollar. And if the dad-clown does a swan, we abort." "He won't do a swan. Nobody does a swan."
created_at
"2026-09-24T14:12:34.774Z"
9/24/2026, 2:12:34 PM
model
"aion-labs/aion-3.5-mini-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1648
moderation_latency
(null)
generation_time
207153
tokens_prompt
75
tokens_completion
9064
native_tokens_prompt
65
native_tokens_completion
10117
native_tokens_completion_images
(null)
native_tokens_reasoning
8232
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
0.0141755
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790259154-B4f0TR82rBopjsGXnrWg"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790259154-oQJOefM8lc42zt8nVISk"
upstream_id
"00c4e95ae6704117908b6ad66a9c0d10"
provider_responses
0
endpoint_id
"60ac3bdc-90ec-4769-9193-2c00b4d79562"
id
"00c4e95ae6704117908b6ad66a9c0d10"
is_byok
false
latency
1648
model_permaslug
"aion-labs/aion-3.5-mini-20260923"
provider_name
"AionLabs"
status
200
total_cost
0.0141755
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
0.49%
Matches word count
n/a
neededClean
false
words
527
7.83%
Dialogue to Total Word Ratio
Ratio: 77.14%, Deviation: 7.14%
neededClean
false
wordsTotal
538
wordsDialogue
415
4.1609%