Kimi K3.1 Pro 是 Kimi K3.1 系列的首款模型，面向长程编程、3D 创作与金融研究等复杂任务。该模型在真实代码库中的问题修复、多步工具调用与端到端项目构建上表现突出，也可处理游戏引擎脚本、三维建模与交互式场景生成。该模型原生支持视觉理解并可接受视频输入，适合图文混合内容的分析与生成。其思考模式常开，允许按需调节推理强度，便于在需要多轮规划、外部工具协作与结果核验的智能体场景中使用。上下文窗口约 100 万 token，可用于大体量代码库、长文档与跨文档核对。该模型提供函数调用与结构化输出支持，可用于对返回格式有严格约束的自动化流程。
入力価格
市場価格：¥26
/ M
出力価格
市場価格：¥136
/ M
キャッシュ読取
市場価格：¥1.3
/ M
キャッシュ書き込み
市場価格：¥26
/ M
入力
テキスト
出力
テキスト
価格は利用ティアとリクエスト条件に応じて変動します
ティア別価格表
| ティア | 入力 | 出力 | キャッシュ読取 | キャッシュ書込 |
|---|---|---|---|---|
base | ¥21.32 | ¥111.52 | ¥1.066 | ¥21.32 |
API
EasyAPI の OpenAI 互換 API でこのモデルを呼び出す、コピーして使えるコード。
Moonshot AI
Coding-focused variant in the Kimi K2 family. Accepts text and image inputs, with reasoning on by default. Suited for long-horizon coding and agentic workflows.
¥4.4828/ M
¥18.6208/ M
¥0.8966/ M
¥5.2727/ M
Moonshot AI
Kimi K3.1 Pro is the first model in the Kimi K3.1 series, built for complex tasks such as long-horizon coding, 3D creation, and financial research. It excels at fixing issues in real codebases, multi-step tool use, and end-to-end project building, and can also handle game engine scripting, 3D modeling, and interactive scene generation. The model natively supports vision understanding and accepts video input, making it suitable for analyzing and generating mixed text-and-image content. Its thinking mode is always on, with reasoning effort adjustable on demand, making it useful for agent scenarios that require multi-turn planning, external tool collaboration, and result verification. It has a context window of about 1 million tokens and can be used for large codebases, long documents, and cross-document verification. The model supports function calling and structured output, and can be used in automated workflows with strict constraints on response format.
¥10.66/ M
¥55.76/ M
¥1.066/ M
¥10.66/ M
DeepSeek
The DeepSeek V4 flagship, with the series' strongest deep reasoning and code generation, suited to whole-codebase analysis, multi-step automation and demanding instructions.
¥1.981/ M
¥5.943/ M
¥0.066/ M
—
DeepSeek
DeepSeek-V4.1-Flash is DeepSeek's current mainstay model; its official API model name is deepseek-flash. It supports a 1M context and up to 384K output, with thinking and non-thinking modes switchable by default, features image understanding, tool use, and JSON Output, and is compatible with the OpenAI and Anthropic APIs.
¥0.42/ M
¥1.68/ M
¥0.0084/ M
—