NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-01-01M1ENDDNWAHJKVYENAQNCK8KX
Run Details
Model:
Grok 4.6 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
1m 27s
Cost:
$0.021648
Provider:
xAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
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 paced the warehouse floor, blueprints of the park crumpled in her grip. "Bill, this kids' party is our golden ticket. Presents stacked to the ceiling, bikes, games, the works. We sneak in, grab everything, and we're rich." Bill chewed his sandwich slowly. "Molly, stealing from children? That's a new low. How do we even get past the parents without looking like creeps?" "We disguise as the party clowns. Big red noses, floppy shoes, honking all the way to the gift table. No one questions the entertainment." "Clowns terrify kids now. We'll start a panic. Parents will pounce on us before we touch a single bow." "Then we pose as extra parents. I'll do the frantic mom look, you the bored dad scrolling his phone. Blend in, wait for the cake, then swipe." "I don't look like a dad. I look like a guy who lives in a van. They'll spot us instantly." Molly waved her hand. "Details later. We create chaos with a fake magician act. You pull scarves, I juggle plates, everyone laughs, we load the loot into a fake piƱata." "I can't juggle or do magic. Last time I tried cards they flew everywhere. Kids will laugh at us, not with us." "That's the comedy gold, Bill. We trip over streamers, presents scatter, we scoop them while dodging the bouncy castle." "Bouncy castle? We'll bounce right into the fence and get stuck. This is amateur hour." "Amateur? We're professionals. We hire a real magician as decoy, he does the tricks, we empty the pile during the applause." Bill shook his head. "Hiring costs money we don't have. And applause? At a kids' party it's screaming and crying." "Perfect cover. Noise drowns us out. We use an ice cream van as getaway, kids run to it, we toss presents in the freezer." "Ice cream van at a birthday? They've got cake already. Parents will think we're competing vendors and kick us out." "Then balloons. Hundreds of them. Release, kids chase, we grab and dash through the park." "Helium's expensive and balloons float away. We'll be left holding empty strings while cops arrive." Molly grinned. "Cops at a party? Unlikely. Just moms with cameras. We jam phones with a cheap gadget, time it for candle blowing." "Gadgets from where? Dollar store junk fails. We'll be filmed anyway." "We recruit a couple kids with candy. Tell them it's a pirate game, they help carry the 'treasure'." "Using kids against their own party? That's monstrous. They'll tattle in seconds." "Not if we make it fun. They'll think it's the best game ever until we're gone." Bill sighed. "This whole thing is a disaster. What if someone gets hurt in the chaos?" "No one gets hurt. Pure slapstick. We slip on punch, chase involves unicycles we 'borrow'." "Unicycles? I can't ride a bike. We'll crash into the cake." "Even better. Cake fight as extra diversion. Frosting everywhere, we escape sticky but loaded." "Sticky and arrested. I say we wait till it's over and pick leftovers." "Leftovers? That's not a heist, that's scavenging. We need the thrill, the comedy of it all going wrong but working." They bickered on, Molly sketching wilder ideas like remote cars dragging small gifts, Bill shooting them down with practical fears of getting caught by a six-year-old's stare. Finally Bill muttered, "Fine, but if we end up in timeout with the kids, it's on you." Molly clapped. "That's the spirit. Operation Present Heist is on."
created_at
"2026-09-01T14:19:19.108Z"
9/1/2026, 2:19:19 PM
model
"x-ai/grok-4.6-20260810"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2706
moderation_latency
(null)
generation_time
86552
tokens_prompt
75
tokens_completion
1508
native_tokens_prompt
267
native_tokens_completion
3583
native_tokens_completion_images
(null)
native_tokens_reasoning
2772
native_tokens_cached
256
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
"completed"
service_tier
"default"
usage
0.021648
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788272359-71m38QXtTbtzQIBF9RBU"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788272359-DvuYTN4N3AXGvlNb2CBy"
upstream_id
"1b10bbdb-7dc9-9891-98a9-f50334a17ccb"
provider_responses
0
endpoint_id
"0d0536e3-7eb1-4acb-8249-c8813365c2d8"
id
"1b10bbdb-7dc9-9891-98a9-f50334a17ccb"
is_byok
false
latency
141
model_permaslug
"x-ai/grok-4.6-20260810"
provider_name
"xAI"
status
200
total_cost
0.021648
cache_discount
0.000384
upstream_inference_cost
0
provider_name
"xAI"
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
577
0.00%
Dialogue to Total Word Ratio
Ratio: 88.95%, Deviation: 18.95%
neededClean
false
wordsTotal
579
wordsDialogue
515
0.0000%