NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HGJS596FSMHDYT2BDZ6TRW
Run Details
Model:
Gemini 3.8 Flash (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
22.6s
Cost:
$0.032878
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
100.0000%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 unrolled the blueprints of young Timmy Higgins’s suburban backyard with grave professional solemnity. Across the booth, Bill adjusted his goggles, which were entirely useless in the diner’s fluorescent lighting. A red crayon cross marked the gift table, nestled deep behind an obstacle course composed of folding chairs, an oscillating sprinkler, and thirty screaming children. "The inflatable bounce castle provides maximum tactical cover," Molly murmured, tapping the vinyl turret with her pen. She detailed their synchronized perimeter breach. At precisely two o’clock, during the birthday anthem, she would deploy a smoke grenade disguised as a glitter cannon. That distraction granted them ninety seconds to secure the pricy drones, boxed video games, and the massive luxury Lego set before the clown could sound any alarm. "Just watch out for hyperactive toddlers hopped up on cheap birthday cake," Bill warned, testing the heavy brass zippers on two empty camouflage duffel bags. A single sticky handprint could easily ruin their forensic integrity, yet the promised loot made the terrifying danger entirely worth the immense risk. They solemnly bumped fists over cold, greasy French fries, fully prepared to strike the suburban backyard without ever showing any mercy when the donkey piñata finally burst open.
created_at
"2026-09-02T16:52:35.123Z"
9/2/2026, 4:52:35 PM
model
"google/gemini-3.8-flash-20260902"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1645
moderation_latency
(null)
generation_time
22521
tokens_prompt
75
tokens_completion
2494
native_tokens_prompt
67
native_tokens_completion
8754
native_tokens_completion_images
0
native_tokens_reasoning
8506
native_tokens_cached
0
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
"default"
usage
0.03287775
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788367955-L5U9l8ZqAVt80GYaBoz9"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788367955-F8osM9Jl9kRp4mdvPzpY"
upstream_id
"U1SYat2qDO_UjMcPg_CyuA0"
provider_responses
0
endpoint_id
"d376f781-516d-42f9-936d-46f19abca6cc"
id
"U1SYat2qDO_UjMcPg_CyuA0"
is_byok
false
latency
1645
model_permaslug
"google/gemini-3.8-flash-20260902"
provider_name
"Google AI Studio"
status
200
total_cost
0.03287775
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
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
200
100.00%
Dialogue to Total Word Ratio
Ratio: 10.00%, Deviation: 0.00%
neededClean
false
wordsTotal
200
wordsDialogue
20
100.0000%