# Kimi Code

Canonical page: https://agentsky.dev/agents/kimi-code

Moonshot's official coding agent, multi-model support.

Kimi Code is Moonshot AI's first-party agentic coding assistant, with Kimi K3 as its primary model and broad cross-provider support for DeepSeek, GPT, and GLM models. It is the only harness purpose-built for the Kimi model family and handles Moonshot's API specifics natively.

## Facts

- **Agent type:** `kimi_code`
- **Default model:** Kimi K3 (`kimi-k3`)
- **Start in the browser:** https://agentsky.dev/playground/enter?agentType=kimi_code&llm=kimi-k3&source=agent-md
- **API reference:** https://agentsky.dev/docs
- **Billing:** Running sessions are billed by compute time and model tokens; parked agents incur no compute charge.

## Models it runs on

USD per 1M tokens.

| Model | Id | Input | Output |
|---|---|---|---|
| Kimi K3 (default) | `kimi-k3` | $3.30 | $16.50 |
| Kimi K3.1 Pro | `kimi-k3.1-pro` | $2.20 | $11.00 |
| Kimi K2.7 Code | `kimi-k2.7-code` | $1.045 | $4.40 |
| DeepSeek V4 Pro | `deepseek-v4-pro` | $0.924 | $2.772 |
| DeepSeek V4.1 Flash | `deepseek-flash` | $0.21 | $0.84 |
| GPT-5.6 Sol | `gpt-5.6-sol` | $3.50 | $21.00 |
| GPT-5.6 Terra | `gpt-5.6-terra` | $1.40 | $8.40 |
| GPT-5.6 Luna | `gpt-5.6-luna` | $0.14 | $0.84 |
| GPT-6 Astra | `gpt-6-astra` | $7.00 | $35.00 |
| GPT-6.1 Sol | `gpt-6.1-sol` | $1.40 | $7.00 |
| GPT-6 Sol | `gpt-6-sol` | $1.40 | $7.00 |
| GPT-6 Luna | `gpt-6-luna` | $0.07 | $0.35 |
| GLM-5.3 | `glm-5.3` | $1.54 | $4.84 |
| GLM-5.3-Flash | `glm-5.3-flash` | $0.165 | $0.55 |
| Grok 4.6 | `grok-4.6` | $2.20 | $6.60 |
| Claude Opus 5.5 | `claude-opus-5-5` | $2.80 | $14.00 |
| Claude Opus 5 | `claude-opus-5` | $3.50 | $17.50 |
| Claude Fable 5.1 | `claude-fable-5-1` | $7.00 | $35.00 |
| Claude Sonnet 5 | `claude-sonnet-5` | $1.40 | $7.00 |
| Claude Haiku 4.5 | `claude-haiku-4-5` | $0.70 | $3.50 |
| Claude Sonnet 5.5 | `claude-sonnet-5-5` | $1.40 | $7.00 |
| Claude Haiku 5.5 | `claude-haiku-5-5` | $0.07 | $0.35 |
| Gemini 4 Argon | `gemini-4-argon` | $4.40 | $22.00 |
| Gemini 3.8 Flash | `gemini-3.8-flash` | $1.65 | $8.25 |
| Gemini 3.5 Flash-Lite | `gemini-3.5-flash-lite` | $0.33 | $2.75 |
| Grok 4.7 | `grok-4.7` | $2.20 | $6.60 |
| Muse Spark 1.3 | `muse-spark-1.3` | $1.375 | $4.675 |
| Muse Spark 1.2 | `muse-spark-1.2` | $1.375 | $4.675 |
| Muse Spark 1.1 | `muse-spark-1.1` | $1.375 | $4.675 |
| Muse Wonder 1.0 | `muse-wonder-1.0` | $4.40 | $22.00 |
| MiniMax-M3 | `minimax-m3` | $0.33 | $1.32 |
| MiniMax-M3.1-Flash | `minimax-m3.1-flash` | $0.165 | $0.66 |

## What it can do

- **Write & edit code:** Long-context coding agent — holds an entire large repo in one pass.
- **Full codebase reads:** Reads and reasons over hundreds of files without truncation.
- **Run tests:** Executes your test suite and fixes failures iteratively.
- **Multi-model:** Kimi K3, DeepSeek V4 Pro, GPT-5.6 Sol, and GLM-5.3.

## Questions

### How do I run Kimi Code in the cloud on AgentSky without setting up my own server or managing a runtime environment?

Create an AgentSky account and start a Kimi Code agent via the API or web UI. AgentSky provisions a persistent cloud computer, installs the harness, and runs your task. Select Kimi K3 (the default) or another compatible model in the same API call.

### Which models can I use with Kimi Code on AgentSky, and is it limited to Kimi models only?

Kimi Code on AgentSky supports Kimi K3, Kimi K3.1 Pro, Kimi K2.7 Code, DeepSeek V4 Pro, DeepSeek V4.1 Flash, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-6 Astra, GPT-6.1 Sol, GPT-6 Sol, GPT-6 Luna, GLM-5.3, GLM-5.3-Flash, Grok 4.6, Claude Opus 5.5, Claude Opus 5, Claude Fable 5.1, Claude Sonnet 5, Claude Haiku 4.5, Claude Sonnet 5.5, Claude Haiku 5.5, Gemini 4 Argon, Gemini 3.8 Flash, Gemini 3.5 Flash-Lite, Grok 4.7, Muse Spark 1.3, Muse Spark 1.2, Muse Spark 1.1, Muse Wonder 1.0, MiniMax-M3, and MiniMax-M3.1-Flash. Kimi K3 is the primary model; the others provide flexibility for cost and provider preference.

### What does it cost to run Kimi Code on AgentSky per task, and how is it billed?

Costs depend on the model selected and token usage per task. DeepSeek V4.1 Flash is the lowest-cost option; GPT-5.6 Sol and Kimi K3 sit at higher per-token rates. Playground coverage for Kimi Code is expanding, so task-level median benchmarks are not yet published. Rates are on the pricing page.

### My Kimi Code agent stops partway through a task and does not finish the requested code changes — what should I look for first?

The most common causes are: a context window overflow when reading many files, a tool call that returned an error Kimi Code treated as terminal, or an ambiguous task description that led to an early stop. Check the task log for the last action before the stop. Scope the task to a single module and provide an explicit success condition in the starting prompt to reduce premature stops.

### If a Kimi Code agent is running a task and the AgentSky cloud machine is restarted or replaced, will it lose the files it was editing?

The session workspace persists across ordinary restarts. If the underlying machine is replaced, AgentSky restores the latest durable recovery point, preserving the working directory and recent file edits. Work at the exact moment of a hard replacement may replay from the last checkpoint. Committing intermediate changes with git adds a durable recovery layer independent of AgentSky's checkpoint system.

### How does running Kimi Code on AgentSky compare to cloning it and running it locally on my own development machine?

Running Kimi Code locally means managing the runtime environment, handling crashes manually, and losing the agent when your laptop closes or network drops. On AgentSky, Kimi Code runs in a persistent cloud computer that auto-recovers, stays alive with the laptop closed, and is reachable via API or channels.