NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-05-10-01KR91ZZWJJVNWYP225WN7TF9S
Run Details
Model:
Gemma 4 26B
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
5.8s
Cost:
$0.000108
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
99.74%
Matches word count
(details)
96.08%
Dialogue to Total Word Ratio
(details)
97.9112%
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
"Is that a clown nose in your pocket, Bill?" Molly whispered, glaring at him through the binoculars. "It’s tactical camouflage!" Bill hissed back. "Nobody suspects the birthday clown." "We are here for the high-end electronics and designer LEGO sets, not to juggle for toddlers." "I’m just saying, the element of surprise is key. If I honk this horn at the right moment, they’ll be too distracted to notice us emptying the gift table." "You are a liability. We sneak in through the patio, grab the loot, and vanish before the cake is cut." "But what about the piñata? It’s a goldmine of premium confectionery." "No piñata heists, Bill! We are professionals, not sugar-addicts." "Fine, fine. But if a kid asks me to make a balloon animal, I’m not promising anything." "Just stay focused. Do you have the heavy-duty gift bags?" "Got 'em. Extra large, reinforced handles, very discreet." "And the getaway tricycle?" "It’s a customized mountain bike with a basket. Very fast. Very stealthy." "If we get caught because you’re riding a tricycle, I am leaving you to explain the clown nose to the parents." "Deal. Now, let’s go. I think the bouncy castle is starting."
created_at
"2026-05-10T13:42:27.482Z"
5/10/2026, 1:42:27 PM
model
"google/gemma-4-26b-a4b-it-20260403"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
96
moderation_latency
(null)
generation_time
5607
tokens_prompt
77
tokens_completion
321
native_tokens_prompt
79
native_tokens_completion
300
native_tokens_completion_images
(null)
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
(null)
usage
0.00010753
router
(null)
provider_responses
0
endpoint_id
"d71c6981-347d-4622-a86b-2c9e5d0f2c6c"
id
"chatcmpl-R6yOkvL9cUH6z4rG88KBgBhi"
is_byok
false
latency
42
model_permaslug
"google/gemma-4-26b-a4b-it-20260403"
provider_name
"DeepInfra"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1778420547-qaSAwUUx6KO0vxcYsxCU"
session_id
(null)
api_type
"completions"
id
"gen-1778420547-ZKVFCQ4o6sedMCvaml5b"
upstream_id
"chatcmpl-R6yOkvL9cUH6z4rG88KBgBhi"
total_cost
0.00010753
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
99.74%
Matches word count
n/a
neededClean
false
words
196
96.08%
Dialogue to Total Word Ratio
Ratio: 94.47%, Deviation: 4.47%
neededClean
false
wordsTotal
199
wordsDialogue
188
97.9112%