K

Kimi

⭐⭐⭐⭐⭐ 4.6/5

200K context window AI perfect for long documents and research

chatbotlong-contextresearchchinese
Visit Website

Developer

Moonshot AI

Released

Oct 2023

Users

20M+

API

Available

📊 Rating Breakdown

Functionality 4.7/5

200K context, file processing, web search, long document handling

Ease of Use 4.6/5

Clean interface, Chinese language optimized, accessible

Performance 4.6/5

Fast processing, excellent long context handling

Value for Money 4.6/5

Generous free tier, competitive pricing, best value in China

Ecosystem 4.4/5

Growing integrations, Moonshot ecosystem

Privacy & Security 4.5/5

Chinese regulations apply, good for local use

Support 4.4/5

Chinese support, documentation, community

Innovation 4.6/5

Long context window, Moonshot's technical strength

✅ Pros

  • Can process entire books at once
  • Great for academic research
  • Generous free tier
  • Excellent file upload support (PDF, Word, etc.)
  • 200K token context window
  • Strong Chinese language capabilities
  • Web search integration
  • Very popular in China

❌ Cons

  • Less known internationally
  • English performance varies
  • Limited integrations compared to US products
  • Requires Chinese account for best access
  • UI primarily in Chinese
  • Customer support limited
  • Limited API features compared to competitors
  • Less third-party tooling

💰 Pricing

$0.03/1K tokens (128K context)

Visit Website

🧠 Available Models

Kimi K2.6

Open Source

Moonshot AI's latest flagship (2026). Open-source, 1T parameters, 32B active (MoE). Supports 262K context. Strong reasoning and coding.

Parameters: 1T total / 32B active (MoE)
Context: 262K tokens
📝 text
✅ Long context ✅ Open source ✅ Strong Chinese ✅ Tool calling
Input: ¥1.10/1M (cached) / ¥6.50/1M (uncached)
Output: ¥27.00/1M tokens
🔌 API Availablevia Moonshot API, NVIDIA NIM

Kimi K2.5

Open Source

Previous-gen model with 262K context. Support for tool calling, JSON mode, and partial mode.

Parameters: 1T total / 32B active (MoE)
Context: 262K tokens
📝 text
✅ Long context ✅ Tool calling ✅ JSON mode ✅ Free credits
Input: ¥3.00/1M tokens
Output: ¥12.00/1M tokens
🔌 API Availablevia Moonshot API, NVIDIA NIM

Kimi K2-Thinking

Open Source

Advanced reasoning variant with chain-of-thought. Excels at complex problems requiring step-by-step analysis.

Parameters: 1T total / 32B active (MoE)
Context: 262K tokens
📝 text
✅ Chain-of-thought ✅ Deep reasoning ✅ Complex problems
Input: ¥4.00/1M tokens
Output: ¥16.00/1M tokens
🔌 API Availablevia Moonshot API

🧠 Available Models

Kimi K2.6

Open Source

Moonshot AI's latest flagship (2026). Open-source, 1T parameters, 32B active (MoE). Supports 262K context. Strong reasoning and coding.

Parameters: 1T total / 32B active (MoE)
Context: 262K tokens
📝 text
✅ Long context ✅ Open source ✅ Strong Chinese ✅ Tool calling
Input: ¥1.10/1M (cached) / ¥6.50/1M (uncached)
Output: ¥27.00/1M tokens
🔌 API Availablevia Moonshot API, NVIDIA NIM

Kimi K2.5

Open Source

Previous-gen model with 262K context. Support for tool calling, JSON mode, and partial mode.

Parameters: 1T total / 32B active (MoE)
Context: 262K tokens
📝 text
✅ Long context ✅ Tool calling ✅ JSON mode ✅ Free credits
Input: ¥3.00/1M tokens
Output: ¥12.00/1M tokens
🔌 API Availablevia Moonshot API, NVIDIA NIM

Kimi K2-Thinking

Open Source

Advanced reasoning variant with chain-of-thought. Excels at complex problems requiring step-by-step analysis.

Parameters: 1T total / 32B active (MoE)
Context: 262K tokens
📝 text
✅ Chain-of-thought ✅ Deep reasoning ✅ Complex problems
Input: ¥4.00/1M tokens
Output: ¥16.00/1M tokens
🔌 API Availablevia Moonshot API

📜 Version History

Oct 2024 v3.0

Kimi API & Platform

Launched Kimi API for developers and platform services for enterprise integration.

Oct 2023 v1.0

Kimi Chat Launch

Moonshot AI launched Kimi Chat in China with an industry-leading 200K context window for long document analysis.

Mar 2024 v2.0

Kimi Pro & Web Search

Released Kimi Pro with enhanced capabilities, integrated web search, and real-time information access.

Jul 2024 v2.5

Long Document Mastery

Optimized long document processing with improved accuracy, multi-document comparison, and summarization.

Jan 2026 v4.5

Agentic Kimi

Introduced autonomous agent capabilities for complex multi-step tasks and workflow automation.

Feb 2025 v3.5

Multi-Modal Kimi

Added image understanding, audio processing, and code execution capabilities.

Dec 2023 v1.5

File Upload & Analysis

Added comprehensive file upload support including PDF, Word, Excel, PowerPoint, and image analysis.

Aug 2025 v4.0

Kimi Enterprise Suite

Enterprise platform with custom knowledge bases, team workspace, and advanced security features.

📄 Subpages

📖

How to Use
💰

Pricing
🔄

Alternatives
🎁

Free
📋

Changelog