// model_detail
active multimodal
Kimi K2.5
by Moonshot AI
Kimi's most versatile model. Native multimodal architecture, vision + text, thinking and non-thinking modes. 256K context.
Context window 256K 256K tokens
Max output 32K 32K tokens
Pricing /1M tok $0.60 in / $3.00 out per 1M tokens (API)
Released Mar 1, 2026
// specifications
Specs & capabilities
API details
- API model ID
kimi-k2.5- Internal ID
kimi-k2-5- Status
- active
- Type
- multimodal
Capabilities
- Vision
- 👁️ Yes
- Function calling
- 🔧 Yes
Pricing
- Input
- $0.60 / 1M tokens
- Output
- $3.00 / 1M tokens
- Source
- First-party API pricing
Lifecycle
- Predecessor
- Kimi K2
// data_confidence
Data notes
⚠️ Knowledge cutoff not listed
// notes
Additional notes
Native multimodal, thinking + non-thinking
// more_from_moonshot-ai
More Moonshot AI models
Kimi K2 Thinking active reasoning
Thinking model based on K2. General agentic and reasoning capabilities, deep reasoning tasks.
Kimi K2 active balanced
MoE model with 1T total / 32B active params. Exceptional coding and agent capabilities. 256K context.
Moonshot v1 128K legacy balanced
Original Moonshot model. 128K context. Superseded by K2 series.
// external_links