NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCWQK8P2FVH8FEG7Q0368VQ
Run Details
Model:
Gemini 3.5 Flash Lite (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
2.7s
Cost:
$0.001043
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
Molly leaned over the blueprint of the municipal community center with a red grease pencil and a scowl. Are you completely insane, Bill? It is a toddler's birthday party. It is a target, Molly, and targets yield loot. We are professionals, not babysitters. The loot in question consists of juice boxes, half a sheet cake, and twenty-four identical plastic excavators. Do not underestimate the purchasing power of upper-middle-class suburban grandparents, okay? Uncle Gary is an executive at a software firm. He does not buy dollar-store toys. Uncle Gary bought little Timmy a pair of orthopedic socks last year because he thought the boy had flat arches. That was a calculated investment. This year, I heard rumors of a remote-controlled helicopter with a high-definition camera. Imagine the surveillance potential. You want to use a children's toy helicopter to case the diamond district next week? I want to steal it because it looks fun to fly into the punch bowl. Focus. We go in through the ventilation shaft disguised as clowns. I am not wearing the giant shoes, Bill. Last time I wore the shoes, I tripped over a Pomeranian and dropped the smoke bombs in the deep fryer. Fine, we go as magicians. You have the doves? I am allergic to birds, Bill. You know this. Every time I get near a pigeon, my sinuses stage a coup. Rabbits then. We pull a bewildered rabbit out of a hat and make our getaway in the bouncy castle. The bouncy castle is stationary. It does not have an ignition, Bill. How do you plan to drive a bouncy castle into traffic? With sheer determination and a very long extension cord. Are you with me or not? I only came because you said there would be mini quiches. There will be quiches. But first, we secure the loot. Hand me the glitter glue. We have a heist to launch.
created_at
"2026-07-25T15:01:38.21Z"
7/25/2026, 3:01:38 PM
model
"google/gemini-3.5-flash-lite-20260721"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
385
moderation_latency
(null)
generation_time
2607
tokens_prompt
75
tokens_completion
461
native_tokens_prompt
67
native_tokens_completion
409
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
0.0010426
router
(null)
provider_responses
0
endpoint_id
"6bd8f433-79e1-416c-b407-1772eb796c9a"
id
"0s9kaorKFdPPjMcPtICkMA"
is_byok
false
latency
385
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-1784991698-fQ4KFesupVGQXnuMMSvz"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784991698-53LLupQxupZPrfLb5nyV"
upstream_id
"0s9kaorKFdPPjMcPtICkMA"
total_cost
0.0010426
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
316
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 70.00%
neededClean
false
wordsTotal
322
wordsDialogue
0
0.0000%