Image preference Cap is the peer rank for this modality — preference arenas and measured boards (VBench / Open ASR / Open Media categories) when they exist. Empty axes stay blank; Cap never copies an overall score into realism or motion.
Open Media Arena evaluations (Text-to-Image, LMArena I2V, TTS Arena, …) feed this portfolio when the catalog worker has matched this slug. Cards below list Cap axes, arena scores, and evidence sources — not the Language Model Matrix.
IMAGE PREFERENCE CAP · QUALITY AXES
Human preference
78.6
CADLE INSIGHTS
Only measured signals
IMAGE CAP
78
#54/122 · 50% confidence
HUMAN PREFERENCE
78.6
Media Cap quality axis
STRONGEST SIGNAL
Human preference
Cap axis 78.6
DESCRIPTION + DETAILS
PROVIDER
Krea
STATUS
ACTIVE
COMMUNITY REVIEWS
OPERATOR NOTES
Quality, latency, cost, and failure modes — verified production observations.
4.5
2 notes indexed
SUBMIT OPERATOR NOTE
Tap a star to rate
Topic:
5 - Frontier benchmark
Stored locally · synced to your profile
I
Inference Operator
Verified
Oct 23, 2026
Tested Krea 2 Medium on high-concurrency tool loops. Token generation remains consistent under sustained context load, with minimal prompt drift on complex JSON schemas.
A
Autonomous Agent Lead
Verified
Oct 19, 2026
Evaluated Krea 2 Medium on multi-turn code synthesis. Code reasoning and diff application succeed on ~88% of test suites. Best paired with strict system prompt guardrails.
MODEL COMPARISON
Metric
Krea 2 Medium
Provider
Krea
Release
Input / 1M
No data yet
Output / 1M
No data yet
Context
No data yet
Cadle Cap
78
Cap board
#54/122
Human preference
78.6
DATA PROVENANCE
Scores may use different scales — check the Benchmarks tab for source links.
Krea 2 Medium