Kimi (Moonshot AI) website screenshot

Kimi (Moonshot AI)

Moonshot AI is a Beijing-based AI lab that develops the Kimi family of long-context multilingual large language models. The consumer assistant is available at kimi.com (and kimi.ai); the developer platform at platform.moonshot.cn (also platform.kimi.com / platform.kimi.ai) exposes an OpenAI-compatible REST API for chat completions, embeddings, file management, fine-tuning, and model listing. Kimi models advertise very long context windows (8K, 32K, 128K, with newer K2.x models pushing toward 256K) and the platform offers multimodal text, image, and video inputs on the flagship Kimi K2.x line. Open-source weights for prior Kimi releases live under the MoonshotAI GitHub organization.

Kimi (Moonshot AI) publishes 1 API on the APIs.io network: Kimi Chat Completions API. Tagged areas include LLM, Long Context, AI, OpenAI Compatible, and Multimodal.

The Kimi (Moonshot AI) catalog on APIs.io includes 1 event-driven AsyncAPI specification and 1 Spectral governance ruleset.

Kimi (Moonshot AI)’s developer surface includes authentication, documentation, GitHub presence, engineering blog, and 6 more developer resources.

41.7/100 thin ▲ 5.7 Agent 54/100 agent ready Full breakdown ↓
scored 2026-07-21 · rubric v0.4
8 APIs 0 Features
LLMLong ContextAIOpenAI CompatibleMultimodalChina

API Rating

API Evangelist API Evangelist Rating How this is scored →
scored 2026-07-21 · rubric v0.4
Composite quality — 41.7/100 · thin
Contract Quality 15.1 / 25
Developer Ergonomics 4.3 / 20
Commercial Clarity 5.8 / 20
Operational Transparency 3.4 / 13
Governance 6.3 / 12
Discoverability 6.8 / 10
Agent readiness — 54/100 · agent ready
Machine-Readable Contract 18 / 18
Agentic Access Contract 15 / 15
MCP Server 0 / 12
Machine-Readable Auth 10 / 10
Idempotency 0 / 9
Stable Error Semantics 0 / 8
Request/Response Examples 0 / 7
Rate-Limit Signaling 7 / 7
Typed Event Surface 6 / 6
Agent Skills 0 / 5
Well-Known Catalog 0 / 4
Consent & Bot Identity 0 / 3
Improve this rating by publishing the missing artifacts — every area above can be raised, and the full rubric is at apis.io/rating/. This rating is computed from github.com/api-evangelist/kimi-moonshot: open an issue to ask a question, or submit a pull request to add artifacts. Want it done for you? Prioritized profiling — $2,500 →

APIs

Kimi Chat Completions API

OpenAI-compatible chat completions endpoint. Accepts a messages array, model identifier (moonshot-v1-8k / 32k / 128k, kimi-k2.5, kimi-k2.6), and standard sampling parameters. Su...

Kimi Embeddings API

OpenAI-compatible embeddings endpoint that returns vector representations of input text for semantic search, clustering, and retrieval-augmented generation workflows.

Kimi Files API

File management endpoint for uploading documents that can be referenced from chat completions (for example, long-document Q&A and the file_id-based context attachment pattern).

Kimi Fine-Tuning API

Fine-tuning jobs endpoint for customizing Moonshot base models on customer-supplied training data. Mirrors the OpenAI fine-tuning surface.

Kimi Models API

Lists models available to the authenticated account and exposes per-model metadata (context window, modality support, pricing tier).

Kimi Tokenizer API

Helper endpoint exposed by the Moonshot platform for counting tokens against a given model's tokenizer prior to submission, useful for managing long-context budgets.

Kimi Assistant (kimi.com)

Consumer-facing AI assistant at kimi.com (also kimi.ai) powered by the Kimi models. Supports long-document upload, web search grounding, and tool use through a chat UI.

Moonshot Open-Weights Releases

Open-weight Kimi model releases (for example prior Kimi K1 / K2 checkpoints) published under the MoonshotAI GitHub organization for research and self-hosted use.

Collections

Pricing Plans

Rate Limits

Kimi Moonshot Rate Limits

2 limits

RATE LIMITS

FinOps

Event Specifications

Kimi (Moonshot AI) Streaming Chat Completions API

AsyncAPI definition for Moonshot AI's Kimi `POST /v1/chat/completions` streaming response channel. Moonshot's chat completions surface is OpenAI-compatible. When the request bod...

ASYNCAPI

API Governance Rules

Kimi (Moonshot AI) API Rules

8 rules · 1 errors 7 warnings

SPECTRAL

Resources

Documentation 1

Reference material describing how the API behaves

Agent Surfaces 1

MCP servers, agent skills, and machine-readable catalogs

Build 1

SDKs, sample code, and the tooling you integrate with

Access & Security 2

Authentication, authorization, and security posture

Company 2

The organization behind the API

Other 3

Properties that don't map to a standard resource type

Source (apis.yml)

apis.yml Raw ↑
opencollection: 1.0.0
info:
  name: Moonshot AI API
  version: 1.0.0
items:
- info:
    name: Models
    type: folder
  items:
  - info:
      name: 列出模型
      type: http
    http:
      method: GET
      url: https://api.moonshot.cn/v1/models
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 列出当前可用的所有模型。
- info:
    name: Billing
    type: folder
  items:
  - info:
      name: 查询余额
      type: http
    http:
      method: GET
      url: https://api.moonshot.cn/v1/users/me/balance
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 查询您在 Kimi 开放平台上的可用余额、代金券余额和现金余额。
- info:
    name: Files
    type: folder
  items:
  - info:
      name: 文件列表
      type: http
    http:
      method: GET
      url: https://api.moonshot.cn/v1/files
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 列出当前用户上传的所有文件。
  - info:
      name: 上传文件
      type: http
    http:
      method: POST
      url: https://api.moonshot.cn/v1/files
      body:
        type: multipart-form
        data:
        - name: file
          type: text
          value: ''
        - name: purpose
          type: text
          value: ''
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 上传文件用于内容提取、图片理解或视频理解。
  - info:
      name: 获取文件信息
      type: http
    http:
      method: GET
      url: https://api.moonshot.cn/v1/files/:file_id
      params:
      - name: file_id
        value: ''
        type: path
        description: 文件标识符
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 获取指定已上传文件的元数据。
  - info:
      name: 删除文件
      type: http
    http:
      method: DELETE
      url: https://api.moonshot.cn/v1/files/:file_id
      params:
      - name: file_id
        value: ''
        type: path
        description: 文件标识符
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 删除一个已上传的文件。
  - info:
      name: 获取文件内容
      type: http
    http:
      method: GET
      url: https://api.moonshot.cn/v1/files/:file_id/content
      params:
      - name: file_id
        value: ''
        type: path
        description: 文件标识符
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 获取以 `file-extract` 用途上传的文件的提取文本内容。
- info:
    name: Chat
    type: folder
  items:
  - info:
      name: 创建聊天补全
      type: http
    http:
      method: POST
      url: https://api.moonshot.cn/v1/chat/completions
      body:
        type: json
        data: '{}'
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 为聊天消息创建补全结果。支持标准聊天、Partial Mode 和 Tool Use(函数调用)。
- info:
    name: Utilities
    type: folder
  items:
  - info:
      name: 估算 Token 数量
      type: http
    http:
      method: POST
      url: https://api.moonshot.cn/v1/tokenizers/estimate-token-count
      body:
        type: json
        data: '{}'
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 估算给定消息和模型所需的 Token 数量。输入结构与聊天补全几乎相同。
- info:
    name: Batch
    type: folder
  items:
  - info:
      name: 列出批处理任务
      type: http
    http:
      method: GET
      url: https://api.moonshot.cn/v1/batches
      params:
      - name: after
        value: ''
        type: query
        description: 分页游标,传入上一页最后一个 batch 的 ID
      - name: limit
        value: ''
        type: query
        description: 每页数量,默认 20
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 列出当前组织的批处理任务。
  - info:
      name: 创建批处理任务
      type: http
    http:
      method: POST
      url: https://api.moonshot.cn/v1/batches
      body:
        type: json
        data: '{}'
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 创建一个批处理任务。需要先通过文件接口上传一个 purpose="batch" 的 JSONL 文件,然后使用返回的 file_id 创建任务。
  - info:
      name: 获取批处理任务详情
      type: http
    http:
      method: GET
      url: https://api.moonshot.cn/v1/batches/:batch_id
      params:
      - name: batch_id
        value: ''
        type: path
        description: 批处理任务的 ID
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 获取指定批处理任务的状态和详细信息。
  - info:
      name: 取消批处理任务
      type: http
    http:
      method: POST
      url: https://api.moonshot.cn/v1/batches/:batch_id/cancel
      params:
      - name: batch_id
        value: ''
        type: path
        description: 批处理任务的 ID
      auth:
        type: bearer
        token: '{{bearerToken}}'
    docs: 取消一个正在进行的批处理任务。取消后,任务状态将先变为 cancelling,最终变为 cancelled。仅 validating、in_progress、finalizing 状态的任务可以取消。
bundled: true