# Kimi K3

> Moonshot AI's flagship open-weights model for software engineering, knowledge work and deep reasoning, with a 1M-token context.

## Key facts

- Model author: Moonshot AI
- Input: $2.8572 per 1M tokens (starting price)
- Output: $14.2858 per 1M tokens (starting price)
- Cached input: $0.2858 per 1M tokens
- Context window: 1,048,576 tokens
- Maximum output: 1,048,576 tokens
- Input modalities: image, text, video
- Output modalities: text
- Capabilities: reasoning, tools
- Released: 2026-07-16
- Model IDs: `jd/kimi-k3`, `moonshotai/kimi-k3`
- Model page: https://www.realrelay.ai/model/moonshot-ai/kimi-k3

## Overview

Kimi K3 is Moonshot AI's flagship model, designed for frontier intelligence scenarios such as software engineering, knowledge work and deep reasoning. It has native visual understanding, always reasons before answering, and ships with open weights. Its 1M-token context window suits long-horizon coding and end-to-end knowledge work.

Kimi K3 has a 1M-token context window. It accepts text, image and video input and returns text output. It was released on July 16, 2026. Its weights are open.

## Capabilities

Kimi K3 supports reasoning and tool calling. It accepts text, image and video as input and produces text as output.

Kimi K3 accepts text, image and video input and supports tool calls, JSON mode, structured output and automatic context caching. Because it always reasons, it suits tasks where depth of thought matters more than a quick reply.

## Pricing

Kimi K3 starts from $2.86 per 1M input tokens and from $14.29 per 1M output tokens. Cache reads start from $0.286 per 1M tokens.

Pricing is tiered, and the price shown is the starting tier. Kimi K3 is available through RealRelay's OpenAI-compatible API.

## Use cases

Use Kimi K3 for long-horizon coding and software engineering, where the 1M-token context window can hold a large repository and tool calls drive edits and tests.

For knowledge work and deep reasoning, it can read long documents together with images or video, reason through them and return structured output that other systems can parse.

## Call it

OpenAI-compatible chat completions:

```bash
curl https://www.realrelay.ai/v1/chat/completions \
  -H "Authorization: Bearer $REALRELAY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "moonshotai/kimi-k3", "messages": [{"role": "user", "content": "Hello"}]}'
```

## FAQ

### What is Kimi K3?

Kimi K3 is Moonshot AI's flagship model, designed for frontier intelligence scenarios such as software engineering, knowledge work and deep reasoning. It supports reasoning and tool calling and accepts text, image and video input.

### How much does Kimi K3 cost?

Kimi K3 starts from $2.86 per 1M input tokens and from $14.29 per 1M output tokens, with cache reads from $0.286 per 1M tokens. Pricing is tiered, and these are the starting-tier prices.

### What is the context window of Kimi K3?

Kimi K3 has a 1M-token context window.

### What inputs and outputs does Kimi K3 support?

Kimi K3 accepts text, image and video input and returns text output.

### Does Kimi K3 support tool calling?

Yes. Kimi K3 supports tool calling, so it can call functions you define and use the results in multi-step tasks. It also supports reasoning.

### When was Kimi K3 released?

Kimi K3 was released on July 16, 2026.

### How do I call Kimi K3 via the API?

Call Kimi K3 through RealRelay's OpenAI-compatible API using the model ID moonshotai/kimi-k3.