KKMoonshot AIReady now

Kimi K3

A published Kimi K3 endpoint for long-context production inference workloads.

Deploy Endpoint Run model

Context length1,048,576

GPU requirementNot published

Concurrency—

API model IDkimi-k3

Input pricing$3.00 / 1M

Output pricing$15.00 / 1M

Cached input pricing$0.30 / 1M

Recommended use cases

Where this model fits.

Supported features

Ready for production.

OpenAI compatible

Scale to zero

API example

Use the interface you already know.

Every published text endpoint uses an OpenAI-compatible request surface.

curl

curl https://api.inferx.com/v1/chat/completions \
  -H "Authorization: Bearer $INFERX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "model": "kimi-k3", "messages": [...] }'

Deployment options

Choose how you run it.

Run now

Open the InferX playground and test the endpoint.

Continue

Deploy endpoint

Create a production endpoint in your InferX tenant.

Continue

Dedicated capacity

Deploy with private networking and reserved infrastructure.

Continue