---
title: "Moonshot AI AI models — API access and skills via RunAPI"
url: "https://runapi.ai/providers/moonshot-ai.md"
canonical: "https://runapi.ai/providers/moonshot-ai.md"
locale: "en"
provider: "Moonshot AI"
line_count: 1
variant_count: 2
---

# Moonshot AI

Moonshot AI's Kimi K2 — 1T-parameter MoE with 256K context and 58.6% SWE-bench Pro, via one RunAPI key.

Moonshot AI builds the Kimi K2 family — 1 trillion total parameters, 32B active per token, 384 experts per layer — optimized for autonomous coding and multi-agent orchestration. kimi-k2.6 reaches 58.6% on SWE-bench Pro and scales Agent Swarm to 300 sub-agents. Both kimi-k2.5 and kimi-k2.6 are available through RunAPI from the OpenAI and Anthropic SDKs.

## Models

- [Kimi](https://runapi.ai/models/kimi.md) — Text, 2 variants · from $0.020

## Variants

| Model | Variant | Billing | Starting price | URL |
|---|---|---|---|---|
| Kimi | `k2.5` | 1K tokens | $0.020 | https://runapi.ai/models/kimi/k2.5.md |
| Kimi | `k2.6` | 1K tokens | $0.020 | https://runapi.ai/models/kimi/k2.6.md |

## LLM API endpoints

Base URL: `https://runapi.ai`

- `POST /v1/chat/completions` — OpenAI compatible
- `POST /v1/messages` — Anthropic compatible

Use the OpenAI or Anthropic SDK with your RunAPI API key.


## FAQ

### Is this an official %{provider} integration?

RunAPI exposes a managed API surface with transparent pricing, capability, and error behavior.

### Do I need a %{provider} account?

No — your RunAPI key is enough for managed access.

### What's the latency overhead from proxying?

Typically under 20 ms. RunAPI keeps the proxy layer close to model execution regions.

### Are images / videos cached?

Generated outputs are stored and retrievable by task ID. Inputs are not cached.

### Can I bring my own key?

Not currently — calls use RunAPI-managed access.