Kimi
Gain production-grade API access to Moonshot AI’s flagship Kimi K3 and K2 model series. Built to process massive context windows with zero loss alongside advanced Context Caching for lower latency and cost, Kimi excels at complex code generation, multi-step problem solving, automated research, and dynamic tool orchestration.
All Models
Kimi K3.1 Pro
Kimi K3.1 Pro is Moonshot AI's premier frontier reasoning model built on an expansive 3-trillion parameter sparse Mixture-of-Experts (MoE) architecture. Designed for advanced enterprise systems and autonomous multi-agent environments, it introduces native 3D spatial reasoning, allowing users to synthesize complex 3D scenes and structural layouts directly via natural language and code. Kimi K3.1 Pro features a massive 1 million token context window, enabling deep multi-turn agentic execution, full-stack software refactoring, and low-hallucination financial modeling.
kimi-k3 2
Kimi K3 is Kimi’s most capable flagship model to date. With 2.8 trillion parameters, it is built on the Kimi Delta Attention (KDA) hybrid linear attention architecture and Attention Residuals technology. It natively supports visual understanding and features a 1-million-token context window. As the world’s first open-source model at the 3-trillion-parameter scale, Kimi K3 is designed for frontier AI use cases, including long-horizon coding, knowledge work, and reasoning.
K3 Custom Edition
Kimi K3 is Kimi’s most capable flagship model to date. With 2.8 trillion parameters, it is built on the Kimi Delta Attention (KDA) hybrid linear attention architecture and Attention Residuals technology. It natively supports visual understanding and features a 1-million-token context window. As the world’s first open-source model at the 3-trillion-parameter scale, Kimi K3 is designed for frontier AI use cases, including long-horizon coding, knowledge work, and reasoning.
Kimi K3
Kimi K3 is Kimi’s most capable flagship model to date. With 2.8 trillion parameters, it is built on the Kimi Delta Attention (KDA) hybrid linear attention architecture and Attention Residuals technology. It natively supports visual understanding and features a 1-million-token context window. As the world’s first open-source model at the 3-trillion-parameter scale, Kimi K3 is designed for frontier AI use cases, including long-horizon coding, knowledge work, and reasoning.
Kimi Models API Pricing Details
| Model | Pricing (USD) | Our Pricing (USD) | Discount | |
|---|---|---|---|---|
| Kimi K3.1 Pro | $2/M Tokens | Start from$1.6/M Tokens | -20% | |
| kimi-k3 3 | $3/M Tokens | Start from$3/M Tokens | — | |
| kimi-k3 2 | $3/M Tokens | Start from$1.8/M Tokens | -40% | |
| K3 Custom Edition | $3/M Tokens | Start from$1.8/M Tokens | -40% | |
| Kimi K3 | $3/M Tokens | Start from$1.8/M Tokens | -40% |