NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ04E7N04F5666ABY7YJ31BE
Run Details
Model:
Xiaomi MIMO v2.5
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
6.9s
Cost:
$0.000976
Provider:
Xiaomi
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 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
So, we need to synchronize our watches. They're plastic, Bill. They don't tick. Then we synchronize our breathing. On the exhale, we breach the perimeter. You mean the garden gate? It's a symbolic perimeter, Molly. Now, the clown. He's the wildcard. He's making balloon animals, not guarding Fort Knox. Never underestimate a man in face paint. You handle the clown. I'll neutralize the juice box station. You're going to attack a table of juice boxes? I'm going to create a diversion. A sticky, red, sugary diversion. It'll draw the guards. What guards? There's just Dave from next door napping on a lawn chair. Exactly! He's the inside man. He's sleeping on the job. Now, the target: the gift table. It's a fortress of wrapping paper. It's four bags of pre-wrapped toys. And one piñata. That's our vault. We'll need a stepladder and a very precise bat. Bill, the piñata is shaped like a donkey. It's filled with candy. A perfect disguise for the real loot. Everyone goes for the candy. No one will suspect we're after a half-used pack of crayons and a wooden puzzle. You're right. This plan is brilliant. Now, about your outfit... What about it? The tiny burglar mask is a bit much. You look like a raccoon trying to rob a preschool. It's called professional attire. Now, prepare the getaway vehicle. The red wagon? She's a classic. Now, on my signal... which will be when the clown honks his nose. We move.
created_at
"2026-04-24T16:16:19.63Z"
4/24/2026, 4:16:19 PM
model
"xiaomi/mimo-v2.5-20260422"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
625
moderation_latency
(null)
generation_time
6748
tokens_prompt
77
tokens_completion
548
native_tokens_prompt
311
native_tokens_completion
467
native_tokens_completion_images
(null)
native_tokens_reasoning
119
native_tokens_cached
256
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"
usage
0.00097648
router
(null)
provider_responses
0
endpoint_id
"2af4bf94-26fa-42b9-bce0-8b11071e105a"
id
"2449426190d542179317885448c5e404"
is_byok
false
latency
625
model_permaslug
"xiaomi/mimo-v2.5-20260422"
provider_name
"Xiaomi"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777047379-nUiVmdyXU1tJoajEXUl5"
session_id
(null)
api_type
"completions"
id
"gen-1777047379-RqJ8wykduhui2I9FdADy"
upstream_id
"2449426190d542179317885448c5e404"
total_cost
0.00097648
cache_discount
0.00008192
upstream_inference_cost
0
provider_name
"Xiaomi"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
246
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 90.00%
neededClean
false
wordsTotal
247
wordsDialogue
0
0.0000%