NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHXPFH5CWKBPTHT0XJ7TPYXA
Run Details
Model:
Mistral Medium 3.1
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
6.1s
Cost:
$0.000624
Provider:
Mistral
Evaluation summary
Score
Evaluator
Details Link
20.19%
Matches word count
(details)
28.90%
Dialogue to Total Word Ratio
(details)
24.5433%
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
Molly adjusted her oversized clown wig and glared at Bill. “You’re telling me the *best* disguise you could come up with was a guy in a dinosaur costume?” Bill flexed his foam T-Rex arms. “Kids love dinosaurs. I blend in. Also, limited arm mobility—perfect alibi if someone asks why I’m not helping with the piñata.” “You look like a drunk lizard. And what’s with the squeaky shoes?” “Authenticity! Ever met a silent dinosaur? No. Because they’re *extinct*.” He stomped, emitting a high-pitched *eeek*. “See? Distraction tactic.” Molly rubbed her temples. “We’re stealing from *babies*, Bill. Do you have even a *shred* of moral flexibility?” “First, they’re *toddlers*—tiny tyrants with sticky fingers and zero impulse control. Second, have you *seen* the price of Legos? We’re doing their parents a favor. Third—” He pulled out a party invitation. “—it says ‘no gifts over $20.’ *Amateurs.* We’ll be in and out before the cake’s even cut.” Molly sighed. “Fine. But if a three-year-old outsmarts you, I’m leaving you in the ball pit.” Bill gasped. “You wouldn’t.” “Try me, *Rex*.”
created_at
"2026-02-20T14:15:24.088Z"
2/20/2026, 2:15:24 PM
model
"mistralai/mistral-medium-3.1"
3/1/2001, 12:00:00 AM
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
712
moderation_latency
(null)
generation_time
5979
tokens_prompt
77
tokens_completion
315
native_tokens_prompt
69
native_tokens_completion
298
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)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.0006236
router
(null)
provider_responses
0
endpoint_id
"13c5002b-8c5b-490c-8a5f-52ec0242804f"
id
"7dab12e95cd34152bb9befaff99f1076"
is_byok
false
latency
712
model_permaslug
"mistralai/mistral-medium-3.1"
3/1/2001, 12:00:00 AM
provider_name
"Mistral"
status
200
api_type
"completions"
id
"gen-1771596924-LurmCRGPusZpFow0am8l"
upstream_id
"7dab12e95cd34152bb9befaff99f1076"
total_cost
0.0006236
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Mistral"
Evaluation details
Result
Evaluator
Details
Meta Data
20.19%
Matches word count
n/a
neededClean
false
words
180
28.90%
Dialogue to Total Word Ratio
Ratio: 79.44%, Deviation: 10.56%
neededClean
false
wordsTotal
180
wordsDialogue
143
24.5433%