Kimi K2.6: Long-Horizon Coding and Multi-Agent Workflows
Kimi’s official long-horizon cases become a staged engineering workflow with reversible checkpoints, tool logs, and acceptance tests.
Kimi K2.6 · Prompts and workflows
Browse executable prompts and workflows by task, content type, and runtime. Each card shows inputs, verification status, and the original source.
Runtimes vary by resource and may include an API, CLI, agent framework, or browser coding environment. Check each card for details.
Editor picks
Kimi’s official long-horizon cases become a staged engineering workflow with reversible checkpoints, tool logs, and acceptance tests.
The official quickstart covers thinking mode, multimodal input, and tool configuration; detail is limited to a reproducible API call.
All resources
Kimi’s official long-horizon cases become a staged engineering workflow with reversible checkpoints, tool logs, and acceptance tests.
The official quickstart covers thinking mode, multimodal input, and tool configuration; detail is limited to a reproducible API call.
Kimi K2.6
After downloading, check the models and features available to your account. API and local-code steps still require their own runtime.