x

Author: dancer

content/docs/02-foundations/02-providers-and-models.mdx

@@ -37,6 +37,7 @@ The AI SDK comes with a wide range of providers that you can use to interact wit
 - [Fireworks Provider](/providers/ai-sdk-providers/fireworks) (`@ai-sdk/fireworks`)
 - [DeepInfra Provider](/providers/ai-sdk-providers/deepinfra) (`@ai-sdk/deepinfra`)
 - [DeepSeek Provider](/providers/ai-sdk-providers/deepseek) (`@ai-sdk/deepseek`)
+- [Moonshot Provider](/providers/ai-sdk-providers/moonshot) (`@ai-sdk/moonshot`)
 - [Cerebras Provider](/providers/ai-sdk-providers/cerebras) (`@ai-sdk/cerebras`)
 - [Groq Provider](/providers/ai-sdk-providers/groq) (`@ai-sdk/groq`)
 - [Perplexity Provider](/providers/ai-sdk-providers/perplexity) (`@ai-sdk/perplexity`)
@@ -137,6 +138,8 @@ Here are the capabilities of popular models:
 | [Google Vertex](/providers/ai-sdk-providers/google-vertex)               | `gemini-1.5-pro`                            | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |
 | [DeepSeek](/providers/ai-sdk-providers/deepseek)                         | `deepseek-chat`                             | <Cross size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |
 | [DeepSeek](/providers/ai-sdk-providers/deepseek)                         | `deepseek-reasoner`                         | <Cross size={18} /> | <Cross size={18} /> | <Cross size={18} /> | <Cross size={18} /> |
+| [Moonshot](/providers/ai-sdk-providers/moonshot)                         | `kimi-k2-base`                              | <Cross size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |
+| [Moonshot](/providers/ai-sdk-providers/moonshot)                         | `kimi-k2-instruct`                          | <Cross size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |
 | [Cerebras](/providers/ai-sdk-providers/cerebras)                         | `llama3.1-8b`                               | <Cross size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |
 | [Cerebras](/providers/ai-sdk-providers/cerebras)                         | `llama3.1-70b`                              | <Cross size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |
 | [Cerebras](/providers/ai-sdk-providers/cerebras)                         | `llama3.3-70b`                              | <Cross size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |

content/providers/01-ai-sdk-providers/61-moonshot.mdx

@@ -0,0 +1,106 @@
+---
+title: Moonshot
+description: Learn how to use Moonshot's models with the AI SDK.
+---
+
+# Moonshot Provider
+
+The [Moonshot](https://moonshot.ai) provider offers access to powerful language models through the Moonshot API, including their Kimi K2 model with advanced reasoning and coding capabilities.
+
+API keys can be obtained from the [Moonshot Platform](https://platform.moonshot.ai).
+
+## Setup
+
+The Moonshot provider is available via the `@ai-sdk/moonshot` module. You can install it with:
+
+<Tabs items={['pnpm', 'npm', 'yarn']}>
+  <Tab>
+    <Snippet text="pnpm add @ai-sdk/moonshot" dark />
+  </Tab>
+  <Tab>
+    <Snippet text="npm install @ai-sdk/moonshot" dark />
+  </Tab>
+  <Tab>
+    <Snippet text="yarn add @ai-sdk/moonshot" dark />
+  </Tab>
+</Tabs>
+
+## Provider Instance
+
+You can import the default provider instance `moonshot` from `@ai-sdk/moonshot`:
+
+```ts
+import { moonshot } from '@ai-sdk/moonshot';
+```
+
+For custom configuration, you can import `createMoonshot` and create a provider instance with your settings:
+
+```ts
+import { createMoonshot } from '@ai-sdk/moonshot';
+
+const moonshot = createMoonshot({
+  apiKey: process.env.MOONSHOT_API_KEY ?? '',
+});
+```
+
+You can use the following optional settings to customize the Moonshot provider instance:
+
+- **baseURL** _string_
+
+  Use a different URL prefix for API calls.
+  The default prefix is `https://api.moonshot.ai/v1`.
+
+- **apiKey** _string_
+
+  API key that is being sent using the `Authorization` header. It defaults to
+  the `MOONSHOT_API_KEY` environment variable.
+
+- **headers** _Record&lt;string,string&gt;_
+
+  Custom headers to include in the requests.
+
+- **fetch** _(input: RequestInfo, init?: RequestInit) => Promise&lt;Response&gt;_
+
+  Custom [fetch](https://developer.mozilla.org/en-US/docs/Web/API/fetch) implementation.
+
+## Language Models
+
+You can create language models using a provider instance:
+
+```ts
+import { moonshot } from '@ai-sdk/moonshot';
+import { generateText } from 'ai';
+
+const { text } = await generateText({
+  model: moonshot('kimi-k2-0711-preview'),
+  prompt: 'Write a vegetarian lasagna recipe for 4 people.',
+});
+```
+
+Moonshot language models can be used in the `streamText` function
+(see [AI SDK Core](/docs/ai-sdk-core)).
+
+You can create Moonshot language models using a provider instance. The first argument is the model ID, e.g. `kimi-k2-0711-preview`:
+
+```ts
+const model = moonshot('kimi-k2-0711-preview');
+```
+
+You can also use the `.languageModel()` and `.chat()` methods:
+
+```ts
+const model = moonshot.languageModel('kimi-k2-0711-preview');
+const model = moonshot.chat('kimi-k2-0711-preview');
+```
+
+## Model Capabilities
+
+| Model                  | Image Input         | Object Generation   | Tool Usage          | Tool Streaming      |
+| ---------------------- | ------------------- | ------------------- | ------------------- | ------------------- |
+| `kimi-k2-0711-preview` | <Cross size={18} /> | <Check size={18} /> | <Check size={18} /> | <Check size={18} /> |
+
+<Note>
+  Please see the [Moonshot docs](https://platform.moonshot.ai/docs) for more
+  details about the available models. You can also pass any available provider
+  model ID as a string if needed.
+</Note>

examples/ai-core/src/generate-text/moonshot-chatbot.ts

@@ -0,0 +1,43 @@
+import { moonshot } from '@ai-sdk/moonshot';
+import { generateText, tool } from 'ai';
+import 'dotenv/config';
+import * as readline from 'node:readline/promises';
+import { z } from 'zod/v4';
+
+const terminal = readline.createInterface({
+  input: process.stdin,
+  output: process.stdout,
+});
+
+async function main() {
+  const messages = [];
+
+  while (true) {
+    const userInput = await terminal.question('You: ');
+    messages.push({ role: 'user', content: userInput });
+
+    const { text } = await generateText({
+      model: moonshot('kimi-k2-0711-preview'),
+      tools: {
+        weather: tool({
+          description: 'Get the weather in a location',
+          inputSchema: z.object({
+            location: z
+              .string()
+              .describe('The location to get the weather for'),
+          }),
+          execute: async ({ location }) => ({
+            location,
+            temperature: 72 + Math.floor(Math.random() * 21) - 10,
+          }),
+        }),
+      },
+      messages,
+    });
+
+    console.log(`Assistant: ${text}`);
+    messages.push({ role: 'assistant', content: text });
+  }
+}
+
+main().catch(console.error);

examples/ai-core/src/generate-text/moonshot-tool-call.ts

@@ -0,0 +1,27 @@
+import { moonshot } from '@ai-sdk/moonshot';
+import { generateText, tool } from 'ai';
+import 'dotenv/config';
+import { z } from 'zod/v4';
+
+async function main() {
+  const { text } = await generateText({
+    model: moonshot('kimi-k2-0711-preview'),
+    tools: {
+      weather: tool({
+        description: 'Get the weather in a location',
+        inputSchema: z.object({
+          location: z.string().describe('The location to get the weather for'),
+        }),
+        execute: async ({ location }) => ({
+          location,
+          temperature: 72 + Math.floor(Math.random() * 21) - 10,
+        }),
+      }),
+    },
+    prompt: 'What is the weather in Tokyo?',
+  });
+
+  console.log(text);
+}
+
+main().catch(console.error);

examples/ai-core/src/generate-text/moonshot.ts

@@ -0,0 +1,15 @@
+import { moonshot } from '@ai-sdk/moonshot';
+import { generateText } from 'ai';
+import 'dotenv/config';
+
+async function main() {
+  const { text } = await generateText({
+    model: moonshot('kimi-k2-0711-preview'),
+    prompt: 'Count from 1 to 3 and say hello.',
+    temperature: 0.6, // Recommended temperature for Kimi K2
+  });
+
+  console.log(text);
+}
+
+main().catch(console.error);

examples/ai-core/src/stream-text/moonshot-tool-call.ts

@@ -0,0 +1,73 @@
+import { moonshot } from '@ai-sdk/moonshot';
+import { streamText, ModelMessage, ToolCallPart, ToolResultPart } from 'ai';
+import 'dotenv/config';
+import { weatherTool } from '../tools/weather-tool';
+
+const messages: ModelMessage[] = [];
+
+async function main() {
+  let toolResponseAvailable = false;
+
+  const result = streamText({
+    model: moonshot('kimi-k2-0711-preview'),
+    maxOutputTokens: 512,
+    tools: {
+      weather: weatherTool,
+    },
+    toolChoice: 'auto',
+    prompt: 'What is the weather in San Francisco?',
+  });
+
+  let fullResponse = '';
+  const toolCalls: ToolCallPart[] = [];
+  const toolResponses: ToolResultPart[] = [];
+
+  for await (const delta of result.fullStream) {
+    switch (delta.type) {
+      case 'text': {
+        fullResponse += delta.text;
+        process.stdout.write(delta.text);
+        break;
+      }
+
+      case 'tool-call': {
+        toolCalls.push(delta);
+
+        process.stdout.write(
+          `\nTool call: '${delta.toolName}' ${JSON.stringify(delta.input)}`,
+        );
+        break;
+      }
+
+      case 'tool-result': {
+        const transformedDelta: ToolResultPart = {
+          ...delta,
+          output: { type: 'json', value: delta.output },
+        };
+        toolResponses.push(transformedDelta);
+
+        process.stdout.write(
+          `\nTool response: '${delta.toolName}' ${JSON.stringify(
+            delta.output,
+          )}`,
+        );
+        break;
+      }
+    }
+  }
+  process.stdout.write('\n\n');
+
+  messages.push({
+    role: 'assistant',
+    content: [{ type: 'text', text: fullResponse }, ...toolCalls],
+  });
+
+  if (toolResponses.length > 0) {
+    messages.push({ role: 'tool', content: toolResponses });
+  }
+
+  toolResponseAvailable = toolCalls.length > 0;
+  console.log('Messages:', messages[0].content);
+}
+
+main().catch(console.error);

examples/ai-core/src/stream-text/moonshot.ts

@@ -0,0 +1,17 @@
+import { moonshot } from '@ai-sdk/moonshot';
+import { streamText } from 'ai';
+import 'dotenv/config';
+
+async function main() {
+  const result = streamText({
+    model: moonshot('kimi-k2-0711-preview'),
+    prompt: "How many r's is there in strawberry?",
+    temperature: 0.6, // Recommended temperature for Kimi K2
+  });
+
+  for await (const textPart of result.textStream) {
+    process.stdout.write(textPart);
+  }
+}
+
+main().catch(console.error);

packages/moonshot/CHANGELOG.md

@@ -0,0 +1,10 @@
+# @ai-sdk/moonshot
+
+## 1.0.0-beta.1
+
+### Major Changes
+
+- Initial release of the Moonshot provider for AI SDK
+- Support for Kimi K2 models (kimi-k2-base and kimi-k2-instruct)
+- OpenAI-compatible API integration
+- Tool calling support for agentic capabilities

packages/moonshot/README.md

@@ -0,0 +1,57 @@
+# AI SDK - Moonshot Provider
+
+The **[Moonshot provider](https://ai-sdk.dev/providers/ai-sdk-providers/moonshot)** for the [AI SDK](https://ai-sdk.dev/docs) contains language model support for the [Moonshot AI](https://api.moonshot.ai) platform.
+
+## Setup
+
+The Moonshot provider is available in the `@ai-sdk/moonshot` module. You can install it with
+
+```bash
+npm i @ai-sdk/moonshot
+```
+
+## Provider Instance
+
+You can import the default provider instance `moonshot` from `@ai-sdk/moonshot`:
+
+```ts
+import { moonshot } from '@ai-sdk/moonshot';
+```
+
+## Example
+
+```ts
+import { moonshot } from '@ai-sdk/moonshot';
+import { generateText } from 'ai';
+
+const { text } = await generateText({
+  model: moonshot('kimi-k2-0711-preview'),
+  prompt: 'Write a JavaScript function that sorts a list:',
+  temperature: 0.6, // Recommended temperature for Kimi K2
+});
+```
+
+## Models
+
+The Moonshot provider supports the following models:
+
+### Generation Models
+
+- `kimi-k2-0711-preview`: Mixture-of-Experts model with exceptional coding and agent capabilities
+
+## Configuration
+
+You can configure the Moonshot provider with your API key and other settings:
+
+```ts
+import { createMoonshot } from '@ai-sdk/moonshot';
+
+const moonshot = createMoonshot({
+  apiKey: 'your-moonshot-api-key', // defaults to process.env.MOONSHOT_API_KEY
+  baseURL: 'https://api.moonshot.ai/v1', // optional, defaults to Moonshot's API
+});
+```
+
+## Documentation
+
+Please check out the **[Moonshot provider](https://ai-sdk.dev/providers/ai-sdk-providers/moonshot)** for more information.