Access Kimi K3 on any platform
A directory of every provider offering the Moonshot AI Kimi K3 model with 1M token context. Find the right inference platform for your project.
# Install & use in 3 lines pip install openai from openai import OpenAI client = OpenAI( base_url="https://api.moonshot.cn/v1", api_key="your-api-key" ) response = client.chat.completions.create( model="kimi-k3", messages=[{"role":"user","content":"Hello"}] )
Major platforms
10 providers with direct Kimi K3 accessMoonshot AI
OfficialFirst-party API with latest features, full 1M-token context, and priority access to model updates.
platform.moonshot.cn →OpenRouter
PartnerUnified API with automatic provider switching, failover, and competitive routing across backends.
openrouter.ai →Together AI
PartnerProduction-grade infrastructure optimized for speed. Supports K3 and K2 model families.
together.ai →SiliconFlow
PartnerNative inference with optimized performance. Also powers OpenRouter backend infrastructure.
siliconflow.cn →Fireworks AI
PartnerHigh-performance inference available through OpenRouter and direct API integration.
fireworks.ai →DeepInfra
PartnerDrop-in OpenAI-compatible API. Simple migration with existing SDKs and tooling.
deepinfra.com →Novita AI
PartnerFlexible platform hosting Kimi alongside open-source and proprietary model catalogs.
novita.ai →Cloudflare Workers AI
PartnerEdge inference on Cloudflare's global network. Integrated with Workers ecosystem.
developers.cloudflare.com →DigitalOcean GenAI
PartnerManaged Kimi hosting on DigitalOcean. Integrated with their developer cloud platform.
digitalocean.com →Weights & Biases
PartnerHosted inference combined with ML experiment tracking and model monitoring tools.
wandb.ai →