NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-12-01KKHZEQ451EHEGHMQJHAWCWHA
Run Details
Model:
Grok 4.20 (Beta)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
2.0s
Cost:
$0.001730
Provider:
xAI
Evaluation summary
Score
Evaluator
Details Link
99.38%
Matches word count
(details)
98.72%
Dialogue to Total Word Ratio
(details)
99.0497%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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 said, “Bill, we hit the bouncy castle at four fifteen sharp. The clown’s doing his balloon animal bit, that’s when the parents zone out completely.” Bill chuckled. “You want me dressed as the disgruntled dad again? Last time the mom with the juice box kept offering me gluten-free cookies.” “Skip the dad costume. We go as rival magicians. I’ll yell that your rabbit is clearly a gerbil and start a fake fight. Chaos equals opportunity.” “While they’re filming us for TikTok, you slip out back and bag the table gifts?” “Exactly. Those little idiots get fifty presents each these days. We’ll clear at least three grand in unopened Lego alone.” Bill rubbed his hands. “What about the birthday kid? He’s gonna cry, Mol.” “Let him cry. Kid’s turning seven. By next year he’ll be stealing his own iPad. We’re just accelerating his moral education.” “You’re all heart,” Bill snorted. “Fine. I’ll disable the balloon arch so the helium tanks roll everywhere. Nothing distracts adults like floating toddlers.” Molly grinned wickedly. “And if anyone asks, we’re from Party Pros. Worst case, we blame the rival company down the street.” “Beautiful. Let’s go ruin some childhoods.”
created_at
"2026-03-12T21:32:45.065Z"
3/12/2026, 9:32:45 PM
model
"x-ai/grok-4.20-beta-20260309"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
310
moderation_latency
(null)
generation_time
1713
tokens_prompt
77
tokens_completion
337
native_tokens_prompt
182
native_tokens_completion
266
native_tokens_completion_images
0
native_tokens_reasoning
0
native_tokens_cached
128
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"completed"
usage
0.0017296
router
(null)
provider_responses
0
endpoint_id
"33f49f24-a0de-4dcf-872d-ca514e83fbae"
id
"5a31cc52-a6d5-963c-6495-5ffa3d40af33"
is_byok
false
latency
296
model_permaslug
"x-ai/grok-4.20-beta-20260309"
provider_name
"xAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1773351165-6T1ASj5rxmXGPeBB5Luz"
upstream_id
"5a31cc52-a6d5-963c-6495-5ffa3d40af33"
total_cost
0.0017296
cache_discount
0.0002304
upstream_inference_cost
0
provider_name
"xAI"
Evaluation details
Result
Evaluator
Details
Meta Data
99.38%
Matches word count
n/a
neededClean
false
words
195
98.72%
Dialogue to Total Word Ratio
Ratio: 93.37%, Deviation: 3.37%
neededClean
false
wordsTotal
196
wordsDialogue
183
99.0497%