NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCTP698ZBETC1NQ96CSAX4N
Run Details
Model:
Gemini 3.5 Flash Lite (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
3.0s
Cost:
$0.001183
Provider:
Google AI Studio
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
Bill squinted at the blue-tinted blueprint, his nose twitching behind thick glasses. Are you mental, Molly? We are talking about maximum security. It is a bouncy castle, Bill, not Fort Knox. They have got roving patrols of sugared-up toddlers, and the perimeter is guarded by an aggressive Pomeranian named Sir Fluffington. We trained for this exact scenario in the sewers of Prague. Prague was a bank, Molly. This is a themed extravaganza featuring Princess Sparkle-Dust and twenty-two screaming seven-year-olds hopped up on fruit punch. All the more reason the element of surprise is entirely on our side. Who expects master cat burglars at a royal princess jubilee? Anyone with eyes, considering you are currently wearing a striped turtleneck and carrying an oversized sack with a large dollar sign painted on it. It is classic iconography, Bill. It sets a mood. Now, synchronized watches. We strike at two-fifteen, right when the clown starts his tragic balloon-animal routine. I refuse to be taken down by a man in size-twenty-four shoes who smells faintly of cabbage and despair. Focus, partner. The primary target is the monolithic stack of brightly wrapped loot sitting directly beneath the piata. The piata is a biological hazard. It is stuffed with sugary shrapnel and swung blindly by a sugar-addled maniac wearing a blindfold. That is why you are wearing the riot helmet, darling. It provides structural integrity and deflects rogue blows from papier-mch unicorns. And what is my exit strategy when the birthday boy realizes his limited-edition mega-mech robot has been replaced by a brick? You leap from the second-story balcony into the inflatable bounce house, roll into the ball pit, and meet me by the getaway tricycle. A tricycle. You traded the getaway van for a child's three-wheeler with streamers on the handlebars. Eco-friendly, highly maneuverable, and frankly, very chic. We are going to rot in a high-security playpen. Not if we execute the maneuver with absolute, unflinching panouflage. Grab your grappling hook and your juice box. It is go time.
created_at
"2026-07-25T14:25:54.993Z"
7/25/2026, 2:25:54 PM
model
"google/gemini-3.5-flash-lite-20260721"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
374
moderation_latency
(null)
generation_time
2987
tokens_prompt
75
tokens_completion
533
native_tokens_prompt
67
native_tokens_completion
465
native_tokens_completion_images
0
native_tokens_reasoning
0
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
0.0011826
router
(null)
provider_responses
0
endpoint_id
"6bd8f433-79e1-416c-b407-1772eb796c9a"
id
"c8dkaoS0A5fVjMcP0a6MuA8"
is_byok
false
latency
374
model_permaslug
"google/gemini-3.5-flash-lite-20260721"
provider_name
"Google AI Studio"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1784989554-YFLG0RilgDPoE8wzItuZ"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784989554-fC8IFACPuagZvwd9D14s"
upstream_id
"c8dkaoS0A5fVjMcP0a6MuA8"
total_cost
0.0011826
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
334
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 70.00%
neededClean
false
wordsTotal
352
wordsDialogue
0
0.0000%