Description
Kimi Code is an AI coding agent from Moonshot AI built for the terminal and IDE, running on the company’s Kimi K2.7 Code model. It reads a codebase, writes and edits files, executes shell commands, searches code, fetches web content, and can spawn subagents to handle parts of a task in parallel. Goal mode lets it work autonomously toward an objective across multiple turns, while Swarm mode runs several agents on the same task in parallel with live progress tracking. Kimi Code speaks both OpenAI-compatible and Anthropic-compatible protocols, so it can also power other agent tools like Claude Code, OpenCode, and Codex through configuration. It’s available through an official CLI, a VS Code extension, and Agent Client Protocol integration for editors like Zed and JetBrains. Kimi Code access is bundled into Kimi’s membership plans, with usage billed separately through the Moonshot Open Platform for API access. It suits developers who want a lower-cost, agentic coding assistant that can run standalone or plug into their existing tools.
Key Features
- Kimi K2.7 Code model — purpose-built coding model with a 262,144 token context window.
- Subagents — dispatch built-in coder, explore, and plan subagents in isolated contexts.
- Goal mode — start an autonomous objective that Kimi Code works toward across multiple turns.
- Swarm mode — run multiple agents on the same task in parallel with live progress tracking.
- Editor integration via ACP — drive a session from Zed, JetBrains, or any Agent Client Protocol client.
- Cross-tool compatibility — power other agents like Claude Code, OpenCode, and Codex through Kimi’s API.
- Background tasks — move long-running commands and subagents to the background and track them in a tasks panel.
How It Works
Kimi Code installs as a CLI or VS Code extension, then connects to a Kimi account with a single /login command rather than a manually managed API key. Inside a terminal session, developers can ask for direct edits, dispatch subagents for exploration and planning, or start Goal mode with a defined objective that Kimi Code pursues across multiple turns until it’s done or hits a decision point. Swarm mode runs several agents on the same objective at once, with rate-limit-aware retries. Long-running commands can be pushed to the background with Ctrl+B and tracked in a tasks panel. Because Kimi Code speaks OpenAI-compatible and Anthropic-compatible protocols, developers can also point existing tools like Claude Code or Codex at Kimi’s endpoint instead of using the official CLI.
Technical Architecture & Overview
- Core Engine: Kimi K2.7 Code, Moonshot AI’s coding-focused model, with a 262,144 token context window; a kimi-k2.7-code-highspeed variant is available for latency-sensitive workflows.
- Deployment: Local CLI and VS Code extension, plus Agent Client Protocol (ACP) support for editors like Zed and JetBrains; works on macOS, Linux, and Windows (via bundled Git Bash).
- API Surface: OpenAI-compatible and Anthropic-compatible API endpoints, documented on the Moonshot Open Platform, letting third-party tools like Claude Code and OpenCode connect to Kimi’s models.
- Known Limits: Devices inactive for over 30 days are automatically unbound for security. Context window is capped at 262,144 tokens on the K2.7 Code models regardless of Kimi membership tier.
Pros & Cons
| Pros | Cons |
|---|---|
| Kimi K2.7 Code API is priced well below comparable Western coding models, per Moonshot’s own published rates | Context window on K2.7 Code is capped at 262,144 tokens, smaller than some competitors’ million-token windows |
| Free Adagio tier available before any paid membership | Devices inactive for 30+ days are automatically unbound and require re-login |
| Subagents and Swarm mode support real parallel work on a single task | Kimi Code for VS Code is only open to new installs for legacy Python CLI users, per Moonshot’s own documentation |
| Speaks OpenAI- and Anthropic-compatible protocols, so it can plug into tools like Claude Code and Codex | Membership and API billing are separate systems, so cost planning requires understanding both |
| Works consistently across macOS, Linux, and Windows | Agent concurrency is capped even on paid tiers, for example 2 concurrent tasks on Moderato |
Pricing
| Plan | Price |
|---|---|
| Adagio | $0/month (6 agent uses, 1 concurrent task) |
| Moderato | $19/month (60 agent uses, 2 concurrent tasks, 25 Agent Swarm uses) |
| Allegretto | $39/month (150 agent uses, 2 concurrent tasks, 50 Agent Swarm uses, Kimi Claw) |
Platform Availability
macOS | Linux | Windows | VS Code | CLI
Best For
Developers | Engineering teams | Cost-conscious coding workflows | Multi-agent automation
Frequently Asked Questions
Is there a free way to try Kimi Code?
Yes. The Adagio tier is free and includes limited agent usage, per Moonshot’s official membership pricing page. Paid tiers start at $19 per month for higher weekly quotas and concurrency.
Can Kimi Code work with other coding agents like Claude Code?
Yes. Kimi Code’s API speaks both OpenAI-compatible and Anthropic-compatible protocols, with official setup guides for connecting tools like Claude Code, OpenCode, and Codex to Kimi’s endpoint instead of their default provider.
What is the context window for Kimi K2.7 Code?
Kimi K2.7 Code supports a 262,144 token context window on both the standard and high-speed model variants, according to Moonshot’s official API pricing page.
How does Kimi Code differ from Claude Code?
Kimi Code runs on Moonshot AI’s own K2.7 Code model and is bundled into Kimi membership plans starting at $19 per month, while Claude Code runs on Anthropic’s Claude models and is included in Claude’s Pro plan starting at $17 to $20 per month.
What is Swarm mode in Kimi Code?
Swarm mode runs multiple agents on the same task in parallel, with live progress tracking and rate-limit-aware retries, according to Moonshot’s official Kimi Code documentation.
Meta Description: Kimi Code pricing, features, and model details for 2026, sourced from Moonshot AI’s official Kimi Code and pricing pages.
Similar Tools in This Directory
Also in this category: Claude Code, terminal-first agentic coding tool | OpenAI Codex, ChatGPT-bundled coding agent | Cursor, AI-native code editor
