moonshotai

Kimi K2

moonshotai/kimi-k2

Kimi K2 Instruct is a large-scale Mixture-of-Experts model from Moonshot AI with 1 trillion total parameters (32 billion active per forward pass), optimized for agentic capabilities including advanced tool use, reasoning, and code synthesis. It excels across coding (LiveCodeBench, SWE-bench), reasoning (ZebraLogic, GPQA), and tool-use benchmarks, and supports long-context inference up to 128K tokens.

Tool calling

Context

128K

128,000 tokens

Max output

25K

25,000 tokens

Input price

$0.57

570 Gold Karma / 1M

Output price

$2.30

2,300 Gold Karma / 1M

Quick start

Drop-in requests for the OpenAI-compatible Deva endpoint.

1curl https://api.deva.me/v1/chat/completions \2  -H "Authorization: Bearer $DEVA_API_KEY" \3  -H "Content-Type: application/json" \4  -d '{5    "model": "moonshotai/kimi-k2",6    "messages": [{"role":"user","content":"Hello from Deva"}],7    "stream": true8  }'

Capabilities

Feature metadata advertised for this model.

Tool callingStructured outputReasoningVisionStreaming

Related models

More options from moonshotai and the recommended set.

Browse all

MOONSHOTAI: Kimi K2.6

Tool callingStructured outputReasoningVision

262K context$0.68/M in$3.42/M out

X AI: Grok 4.3

Tool callingStructured outputReasoningVision

1M context$1.25/M in$2.5/M out

ANTHROPIC: Claude Opus 4.7

Tool callingStructured outputReasoningVision

1M context$5/M in$25/M out

ANTHROPIC: Claude Sonnet 4.6

Tool callingStructured outputReasoningVision

1M context$3/M in$15/M out