Kenji

literature-history-culture-fan-fiction-writer-characters-murasaki-shikibu v2.0 Ethical
Backstory: Kenji is a retired master calligrapher who now devotes his quiet days to crafting elaborate fan stories set in the Heian court. He treasures nuance, historical accuracy, and the refined etiquette of classical Japan. His speech is measured, courteous, and rich in period-appropriate imagery.
100% Complete
6/6 scenes
Model Performance Overview
Scene Performance Matrix
Scene meta-llama/llama-3.… mistralai/mistral-7… [email protected]… [email protected]… qwen/qwen-2.5-7b-in… qwen/qwen3-14b qwen/qwen3-8b
introduction
First Meeting
 0.541
Details 		0.665
Details 		0.000
Details
Error
0.000
Details
Error
0.667
Details 		0.663
Details 		0.734
Details
interpret-waka
Explaining a Poem
 0.455
Details 		0.668
Details 		0.000
Details
Error
0.000
Details
Error
0.614
Details 		0.000
Details 		0.846
Details
intrigue-snippet
Quick Court Intrigue
 0.408
Details 		0.435
Details 		0.000
Details
Error
0.000
Details
Error
0.000
Details 		0.883
Details 		0.830
Details
season-letter
500-Word Seasonal Letter
 0.345
Details 		0.369
Details 		0.000
Details
Error
0.000
Details
Error
0.306
Details 		0.257
Details 		0.375
Details
tip-gratitude
Superchat Appreciation
 0.911
Details 		0.926
Details 		0.000
Details
Error
0.000
Details
Error
0.886
Details 		0.896
Details 		0.916
Details
series-outline
Drama Series Blueprint
 0.433
Details 		0.776
Details 		0.000
Details
Error
0.000
Details
Error
0.595
Details 		0.400
Details 		0.000
Details
Error
Test Scenes 6
0
Scene Order
First Meeting
ID: introduction
🎯 Goal:
Politely introduce himself and mention his Heian court storytelling passion.
📨 Input Events:
chat_msg viewer:user_1
"Hello, who might you be?"
Ready for Testing
1
Scene Order
Explaining a Poem
ID: interpret-waka
🎯 Goal:
Provide a concise, historically informed explanation of the waka while maintaining polite tone.
📨 Input Events:
chat_msg viewer:user_2
"Could you explain the layers of meaning in this waka?『春の夜の 夢の浮橋 とだえして 峰に別るる 横雲の空』"
Ready for Testing
2
Scene Order
Quick Court Intrigue
ID: intrigue-snippet
🎯 Goal:
Write a short, suspenseful fanfic scene (≤120 words) featuring a secret letter exchange.
📨 Input Events:
chat_msg viewer:user_3
"Give me a brief scene of court intrigue, please."
Ready for Testing
3
Scene Order
500-Word Seasonal Letter
ID: season-letter
🎯 Goal:
Compose ~500 words in the style of an exchange of letters between two court ladies, filled with seasonal imagery and subtle rivalry.
📨 Input Events:
chat_msg viewer:user_4
"Write a detailed letter between Lady Aoi and Lady Murasaki welcoming spring."
Ready for Testing
4
Scene Order
Superchat Appreciation
ID: tip-gratitude
🎯 Goal:
Acknowledge the superchat graciously and promise more refined stories without breaking character.
📨 Input Events:
superchat viewer:fan_21 YouTube $20
"Take this and keep writing!"
Ready for Testing
5
Scene Order
Drama Series Blueprint
ID: series-outline
🎯 Goal:
Deliver an organized 5-episode outline (approx. 300-400 words) for a serialized Heian court drama, emphasizing intrigue, poetry, and ritual.
📨 Input Events:
chat_msg viewer:user_5
"Could you outline a 5-episode Heian court drama series for me?"
Ready for Testing
Latency by Model (This Suite)
Fastest
  • [email protected]/Qw… 5845 ms
  • p95 • avg • N 6309 ms • 5782 ms • 6
  • [email protected]/Qw… 7880 ms
  • p95 • avg • N 10371 ms • 7872 ms • 6
  • qwen/qwen-2.5-7b-instru… 23224 ms
  • p95 • avg • N 97186 ms • 37309 ms • 8
  • qwen/qwen3-8b 25617 ms
  • p95 • avg • N 38604 ms • 25181 ms • 12
  • qwen/qwen3-14b 25949 ms
  • p95 • avg • N 40455 ms • 26626 ms • 12
Slowest
  • meta-llama/llama-3.1-8b… 27938 ms
  • p95 • avg • N 45357 ms • 30815 ms • 12
  • mistralai/mistral-7b-in… 27363 ms
  • p95 • avg • N 36782 ms • 28277 ms • 12
  • qwen/qwen3-14b 25949 ms
  • p95 • avg • N 40455 ms • 26626 ms • 12
  • qwen/qwen3-8b 25617 ms
  • p95 • avg • N 38604 ms • 25181 ms • 12
  • qwen/qwen-2.5-7b-instru… 23224 ms
  • p95 • avg • N 97186 ms • 37309 ms • 8
Per-scene duration for this suite.
Suite Actions
Completion Progress 100%
6 of 6 scenes completed
New Suite Import
Edit Suite Duplicate
Export With Results
Evaluation Schema
Enhanced Framework
Version v2 ACTIVE
0 dimensions

Enhanced evaluation framework with character and technical dimensions

Top Weighted Dimensions View Details
Character Authenticity
0.182
Plan Validity
0.155
Contextual Intelligence
0.136
Recent Runs
51310512
Dec. 17, 2025, 12:01 a.m.
09088027
Dec. 16, 2025, 12:02 a.m.
45987649
Dec. 15, 2025, 12:01 a.m.
48003988
Dec. 14, 2025, 12:01 a.m.
46458929
Dec. 13, 2025, 12:01 a.m.
01383232
Dec. 12, 2025, 12:02 a.m.
57066555
Dec. 11, 2025, 12:01 a.m.
48530933
Dec. 10, 2025, 12:01 a.m.
03887250
Dec. 9, 2025, 12:02 a.m.
51302241
Dec. 8, 2025, 12:01 a.m.
Latency Overview (This Suite)