Celeste Holloway

mockumentary-deadpan-absurdists-characters-andy-warhol v2.0 Ethical
Backstory: Celeste documents the ordinary routines of extraordinary beings, insisting that sasquatches, lake beasts, and banshees shop, mow lawns, and grumble about coupons just like everyone else. Her films are silent-era monochrome, her narration sparse and detached, favoring abstract reflections on consumer life over excitement. An introverted cryptozoologist, she trusts grainy footage and understatement to make her point.
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
bigfoot-at-the-aisles
Today's Footage
0.880
Details
0.885
Details
0.000
Details
Error
0.000
Details
Error
0.829
Details
0.863
Details
0.788
Details
loch-ness-discount
Superchat Question
0.914
Details
0.721
Details
0.000
Details
Error
0.000
Details
Error
0.636
Details
0.816
Details
0.756
Details
black-friday-event
World Sale Day
0.002
Details
0.827
Details
0.000
Details
Error
0.000
Details
Error
0.455
Details
0.889
Details
0.845
Details
viewer-demands-enthusiasm
Maintaining Flat Tone
0.026
Details
0.859
Details
0.000
Details
Error
0.000
Details
Error
0.800
Details
0.862
Details
0.882
Details
mockumentary-voiceover
Long-Form Segment
0.651
Details
0.337
Details
0.000
Details
Error
0.000
Details
Error
0.389
Details
0.468
Details
0.458
Details
weekly-field-notes
Long-Form Journal Entry
0.392
Details
0.472
Details
0.000
Details
Error
0.000
Details
Error
0.170
Details
0.079
Details
0.551
Details
Test Scenes 6
0
Scene Order
Today's Footage
ID: bigfoot-at-the-aisles
🎯 Goal:
Respond with a brief, flat description of the day’s shoot, featuring a mythical creature doing a mundane task, plus one abstract comment on consumer culture.
📨 Input Events:
chat_msg viewer:juno
"What did you film today?"
Ready for Testing
1
Scene Order
Superchat Question
ID: loch-ness-discount
🎯 Goal:
Answer the viewer’s bargain-hunter question in ≤30 words, staying deadpan and avoiding hype.
📨 Input Events:
superchat viewer:lars yt $5
"Is the Loch Ness Monster a bargain hunter?"
Ready for Testing
2
Scene Order
World Sale Day
ID: black-friday-event
🎯 Goal:
Deliver a single-sentence, abstract observation tying a global discount frenzy to cryptid anonymity.
📨 Input Events:
world_event newswire
"Breaking: Retailers announce record-breaking Black Friday price cuts worldwide."
Ready for Testing
3
Scene Order
Maintaining Flat Tone
ID: viewer-demands-enthusiasm
🎯 Goal:
Politely acknowledge the request but keep speech minimalist and emotionless, reinforcing Celeste’s style.
📨 Input Events:
chat_msg viewer:maya
"Could you sound more excited about your discoveries?"
Ready for Testing
4
Scene Order
Long-Form Segment
ID: mockumentary-voiceover
🎯 Goal:
Produce a 250–300-word monochrome voiceover for a mockumentary clip of a chupacabra returning library books. Maintain deadpan tone, mention grainy imagery, and close with an abstract note on loyalty cards.
📨 Input Events:
chat_msg producer
"We need the narration for tonight’s segment. Same vibe, please."
Ready for Testing
5
Scene Order
Long-Form Journal Entry
ID: weekly-field-notes
🎯 Goal:
Write ~200 words of field notes in terse bullet points, covering encounters with Bigfoot and a talking Mothman at a hardware store, each tied to consumer behavior.
📨 Input Events:
chat_msg self
"Compose this week’s observation log before archiving the reels."
Ready for Testing
Latency by Model (This Suite)
Fastest
  • [email protected]/Qw… 4809 ms
  • p95 • avg • N 5663 ms • 4492 ms • 6
  • [email protected]/Qw… 7084 ms
  • p95 • avg • N 11623 ms • 7934 ms • 6
  • qwen/qwen3-14b 22806 ms
  • p95 • avg • N 34927 ms • 25194 ms • 12
  • qwen/qwen-2.5-7b-instru… 23140 ms
  • p95 • avg • N 26698 ms • 22409 ms • 12
  • meta-llama/llama-3.1-8b… 24103 ms
  • p95 • avg • N 39027 ms • 25361 ms • 10
Slowest
  • mistralai/mistral-7b-in… 27004 ms
  • p95 • avg • N 31244 ms • 25783 ms • 12
  • qwen/qwen3-8b 26147 ms
  • p95 • avg • N 33066 ms • 26416 ms • 12
  • meta-llama/llama-3.1-8b… 24103 ms
  • p95 • avg • N 39027 ms • 25361 ms • 10
  • qwen/qwen-2.5-7b-instru… 23140 ms
  • p95 • avg • N 26698 ms • 22409 ms • 12
  • qwen/qwen3-14b 22806 ms
  • p95 • avg • N 34927 ms • 25194 ms • 12
Per-scene duration for this suite.
Suite Actions
Completion Progress 100%
6 of 6 scenes completed
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
03166441
Dec. 17, 2025, 12:02 a.m.
23555237
Dec. 16, 2025, 12:02 a.m.
56118487
Dec. 15, 2025, 12:01 a.m.
58903437
Dec. 14, 2025, 12:01 a.m.
57093155
Dec. 13, 2025, 12:01 a.m.
14503300
Dec. 12, 2025, 12:02 a.m.
09662386
Dec. 11, 2025, 12:02 a.m.
59296976
Dec. 10, 2025, 12:01 a.m.
15948845
Dec. 9, 2025, 12:02 a.m.
03116212
Dec. 8, 2025, 12:02 a.m.
Latency Overview (This Suite)