Kimi K3 resource library

Guides for understanding, testing, and building with Kimi K3.

Start with the model, then move into API implementation, token planning, coding agents, memberships, repositories, and open-model verification. Every guide distinguishes official Kimi facts from this site’s independent product layer.

Choose a starting point

New evaluators should begin with “What is Kimi K3?” and the context guide. Application developers should use the API implementation guide. Engineering teams choosing a coding client should continue through Kimi Code, CodePlan, OpenCode, and the coding workflow guide before granting an agent repository or terminal access.

04 guides

Model fundamentals

Understand what Kimi K3 is, how its long context works, what open-model claims mean, and where to verify official repositories and artifacts.

01 guides

API implementation

Move from a first request to authenticated, observable production use while keeping official Kimi Platform access separate from this independent gateway.

05 guides

Coding tools and workflows

Choose and configure Kimi coding surfaces, agent permissions, repository context, membership, quotas, and review practices.

Long-horizon engineering

Kimi K3 for Coding: Workflows, Context, and Evaluation

Kimi K3 is designed for long-horizon coding: work that extends beyond a single completion into repository exploration, planning, tool calls, implementation, and verification. The best results come from combining model capability with disciplined engineering controls.

Read guide
Agentic coding workflow

Kimi Code Guide: CLI, Models, Plans, and Tool Use

Kimi Code is Moonshot AI’s coding-agent environment for terminal and supported editor workflows. It combines models such as Kimi K3 with repository tools, approvals, sessions, configuration, and integrations for third-party coding clients.

Read guide
Membership and usage

Kimi CodePlan Explained: Membership, Quotas, API Billing, and Tools

Kimi CodePlan searches often mix together Kimi Code membership, coding-agent quotas, Kimi Platform API charges, and independent-service credits. This guide separates those systems and explains how to evaluate a plan by completed engineering work rather than an abstract request count.

Read guide
Terminal coding setup

Use Kimi K3 in OpenCode: Setup and Workflow Guide

OpenCode is a terminal-based coding agent that can use Kimi K3 for model inference. This guide explains the integration boundary, a safe setup process, context and reasoning controls, validation, and the operational details that matter after the first successful prompt.

Read guide
Kimi, OpenClaw, and ClawBot

Kimi Claw and ClawBot: What They Are and How They Connect

Kimi Claw brings Kimi-assisted workflows into an OpenClaw-based environment and can connect to communication channels or device automation. “ClawBot” commonly refers to a bot entry inside a channel, not to the Kimi K3 model itself.

Read guide