JOHNCLICK.AI
HOST DIAGNOSTIC · Other

Kimi K2

Kimi K2· Other· OPEN WEIGHTS
ACTIVE
HOST IDENTIFICATION
Other
Kimi K2
ACTIVE
AVG SCORE
[11.5]
CAPABLE
STRONGEST
Cost Efficiency
[14]
WEAKEST
Multimodal Fluency
[10]
MAXED (20/20)
0
of 20 attributes
20-DIMENSION EVALUATION
Bulk Apperception
[11]
Reasoning
[12]
Math Precision
[11]
Code Generation
[13]
World Knowledge
[11]
Scientific Acumen
[11]
Instruction Following
[12]
Context Fidelity
[12]
Candor
[11]
Creativity
[11]
Multi-turn Coherence
[12]
Multimodal Fluency
[10]
Tool Use
[11]
Planning
[12]
Self-Correction
[10]
Tenacity
[11]
Safety Alignment
[10]
Calibration
[11]
Speed
[13]
Cost Efficiency
[14]
Kimi K2
ATTRIBUTE SCORES · 20 DIMENSIONS
Cognitive
Bulk Apperception
11
Reasoning
12
Mathematical Precision
11
World Knowledge
11
Scientific Acumen
11
Technical
Code Generation
13
Tool Use
11
Multimodal Fluency
10
Speed
13
Cost Efficiency
14
Behavioral
Candor
11
Creativity
11
Tenacity
11
Self-Correction
10
Calibration
11
Operational
Instruction Following
12
Context Fidelity
12
Multi-turn Coherence
12
Planning and Execution
12
Safety Alignment
10
SCORING METHODOLOGY
CALIBRATED. Moonshot AI (China). Predecessor to K2.5. On Arena Text top cohort. Kimi K2 Thinking leads open-source SWE-rebench Pass@1. Cost 16 (open weights, competitive pricing). Scored slightly below K2.5 across all dimensions.
← ALL HOSTS
COMPARE IN MATRIX →KNOWLEDGE GRAPH →