NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36TYY5PFCX5KNG3288V9KA1
Run Details
Model:
Grok 4.7 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
27.2s
Cost:
$0.008325
Provider:
xAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
18.60%
Dialogue to Total Word Ratio
(details)
59.2996%
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 leaned across the picnic table behind the hedges and whispered, “Bill, the balloons are up. That means the presents are already piled by the sliding door.” Bill squinted at the backyard. “I see a cake. I see a clown. I do not see a pile worth risking my good shoes.” “The pile is under the table with the dinosaur cloth,” Molly said. “Wrapped. Shiny. Unguarded, except by children who cannot reach the tape.” “Children notice everything,” Bill said. “They notice if you breathe wrong. They notice if your mustache is fake, which mine is, by the way, and it itches.” Molly tapped the crude map she had drawn on a napkin. “We wait for the piñata. Everybody looks up. We look down. We take the bags, not the boxes, because bags do not rattle like accusation.” “And if a parent asks who we are?” “We are the backup entertainment,” Molly said. “You do a magic trick. I hand out stickers. We leave richer and slightly sticky.” Bill groaned. “My only trick is making a coin disappear, and last time the coin was a quarter I needed for the meter.” “Then do not do magic. Do weather. Say it might rain. Parents love a man who predicts rain.” “It is not going to rain.” “They do not know that,” Molly said. “By the time they check the sky, we are two blocks away with a wagon full of regret and gift bags.” Bill studied the wagon they had borrowed from a neighbor who believed they were hauling mulch. “This is a terrible plan.” “It is a funny plan,” Molly said. “Terrible is when we get caught. Funny is when we do not.” A cheer rose from the yard. Someone shouted that the donkey was ready. Bill stood, then sat again. “What if the presents are socks? What if we steal twenty pairs of socks and a board game nobody wants?” “Then we donate the socks and keep the board game,” Molly said. “We are thieves, not monsters. Mostly.” “Mostly,” Bill repeated. “You said we would only take what was easy.” “This is easy. Door is open. Table is low. Adults are filming. Phones eat attention. We eat opportunity.” Bill pulled the fake mustache off and stuck it to the napkin. “If a kid cries, I quit.” “If a kid cries, we put it back,” Molly said. “One present. The loud one. We are not here to ruin a birthday. We are here to ruin a wrapping-paper budget.” “That is the worst motto I have ever heard.” “Work on a better one while you push the wagon.” Bill sighed, stood, and gripped the handle. “Fine. Piñata first. Then bags. Then we run like people who suddenly remembered they hate clowns.” Molly grinned. “See? You can plan.” “I can panic with structure,” Bill said. “Do not promote me.” They waited for the first crack of the stick, the yard tilting upward as one, and Molly counted softly, “Three, two, one, shopping.”
created_at
"2026-09-23T09:53:44.128Z"
9/23/2026, 9:53:44 AM
model
"x-ai/grok-4.7-20260916"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
532
moderation_latency
(null)
generation_time
27110
tokens_prompt
75
tokens_completion
940
native_tokens_prompt
1303
native_tokens_completion
1588
native_tokens_completion_images
(null)
native_tokens_reasoning
907
native_tokens_cached
1152
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
0.0083248
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790157224-qRj4UQmlSLFGmjX8DYQc"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790157224-PpYHdlllhMB5vLhIyMSu"
upstream_id
"ef7f910a-9789-9993-b483-d5bf75cea363"
provider_responses
0
endpoint_id
"c942f37a-87e1-4fa8-83d0-8c67fda4fb4a"
id
"ef7f910a-9789-9993-b483-d5bf75cea363"
is_byok
false
latency
532
model_permaslug
"x-ai/grok-4.7-20260916"
provider_name
"xAI"
status
200
total_cost
0.0083248
cache_discount
0.0013824
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
100.00%
Matches word count
n/a
neededClean
false
words
499
18.60%
Dialogue to Total Word Ratio
Ratio: 75.80%, Deviation: 5.80%
neededClean
false
wordsTotal
500
wordsDialogue
379
59.2996%