frontierreasoning

Moonshot AI (Kimi)

Pioneering long-context and long-thinking AI through the Kimi model family.

3
Models tracked
2023
Founded
Kimi K1.5
Flagship model
Beijing
HQ
Visit kimi.aiDocumentation

About

About Moonshot AI (Kimi)

Moonshot AI (月之暗面), founded in Beijing in 2023, made waves by launching Kimi with one of the first 1M+ token context windows available via API. Kimi K1.5 is a multimodal long-thinking model trained with extended chain-of-thought reasoning that achieves competitive results on MATH and coding benchmarks. The moonshot-v1 API serves millions of users in China with strong bilingual English/Chinese capabilities.

kimi.ai

Key highlights

  • Kimi pioneered 1M+ token context windows via public API in 2023

  • Kimi K1.5 competes with OpenAI o1 on MATH 500 with multimodal support

  • Millions of active users via the Kimi Chat app in China

  • Strong bilingual English/Chinese across all Kimi models

  • Developer API available via the Moonshot platform

Models

3 models tracked

Context windows, pricing, capabilities, and API identifiers — everything you need to choose the right model for your workload.

Flagship

Kimi K1.5

2025

128Kcontext
TextVisionCodeReasoning

A multimodal long-thinking model trained to reason step-by-step before answering. Competitive with o1 on MATH 500 and LiveCodeBench. Supports text, images, and 128K context.

  • Long chain-of-thought reasoning
  • Vision understanding
  • Math & coding benchmarks
  • Bilingual
kimi-k1-5

moonshot-v1-128k

2024

128Kcontext
TextCode

General-purpose model with 128K context for long document processing and extended conversations. Widely used in China for enterprise applications.

  • 128K context
  • Strong bilingual
  • Reliable instruction following
  • Production ready
In: $1.30 / M tokens
Out: $1.30 / M tokens
moonshot-v1-128k

moonshot-v1-32k

2024

32Kcontext
TextCode

Faster and more cost-efficient for standard-length document tasks. The 32K context handles most real-world prompts at lower latency.

  • Fast inference
  • Lower cost
  • Standard document tasks
  • High throughput
In: $0.40 / M tokens
Out: $0.40 / M tokens
moonshot-v1-32k

Use Cases

What Moonshot AI (Kimi) is best for

Each provider has specialisms shaped by their architecture decisions, training data, and API design. Here is where Moonshot AI (Kimi) performs best.

Ultra-long document processing
Mathematical reasoning tasks
Bilingual business applications
Extended conversation systems
Research literature synthesis

Get started

Start building with Moonshot AI (Kimi)

Visit the official Moonshot AI (Kimi) platform to access API keys, documentation, and pricing details.

Visit kimi.aiRead the docsBack to all providers