Model details
Kimi-K2-Instruct-0905 is a large-scale Mixture-of-Experts language model with the listed price trillion total parameters and 32 billion activated per forward pass, developed by Moonshot AI as part of the K2 series. The 0905 iteration, released in September 2025, extends the earlier 0711 build with a longer context window reaching 262K tokens, enabling more complex multi-turn interactions without prompting constraints. Designed for agentic use cases, it emphasizes coding reliability, front-end development, and structured tool calling—positioning it as a practical choice for developers building autonomous workflows and coding assistants.
The 0905 update builds on the K2 lineage with refined agentic coding capabilities and improved reasoning across long contexts. Multiple inference providers offer the model with prompt caching for cost efficiency, and LoRA-based fine-tuning allows task-specific customization. Its blend of massive MoE scale, extended context, and coding-focused strengths makes it well-suited for long-horizon agentic tasks and production-grade deployments.
Transparent token rates
Rates are shown per one million tokens. Combined means one million input plus one million output tokens.
Hugging Face
Kimi K2 0905 is the September update of Kimi K2 0711. It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32 billion active per forward pass. It supports long-context inference up to 256k tokens, extended from the previous 128k. This update im
Hugging Face
Kimi K2 0905 is an updated version of Kimi K2, a state-of-the-art mixture-of-experts (MoE) language model with 32 billion activated parameters and 1 trillion total parameters. Kimi K2 0905 has improved coding abilities, a longer context window, and agentic tool use, and a longer (262K) context window.
