LLM ReferenceLLM Reference

Kimi K2 Thinking

Proprietary

About

Extended thinking variant of Kimi K2 with native reasoning capabilities. 256K context.

Capabilities

VisionMultimodalReasoningFunction CallingTool UseStructured OutputsCode Execution

Providers(5)

Compare all →
ProviderInput (per 1M)Output (per 1M)Type
Fireworks AI$0.6$2.5Serverless
GCP Vertex AI$0.6$2.5Serverless
NVIDIA NIMServerless
AWS BedrockServerless
OpenRouter$0.6$2.5Serverless

Rankings

Specifications

FamilyKimi K2
Released2025-01-01
Context256K
ArchitectureDecoder Only
Specializationreasoning
Trainingpretraining

Created by

Lossless long-context AI innovation

Beijing, China
Founded 2023
Website