feat: improve base tool type (#709)

Co-authored-by: Alex Yang <himself65@outlook.com>

.changeset/config.json

@@ -1,7 +1,7 @@
 {
   "$schema": "https://unpkg.com/@changesets/config@2.3.1/schema.json",
   "changelog": "@changesets/cli/changelog",
-  "commit": true,
+  "commit": false,
   "fixed": [],
   "linked": [],
   "access": "public",

.changeset/nasty-buttons-confess.md

@@ -0,0 +1,5 @@
+---
+"llamaindex": patch
+---
+
+Fix agent streaming with new OpenAI models

.cjs.swcrc

@@ -0,0 +1,16 @@
+{
+  "jsc": {
+    "parser": {
+      "syntax": "typescript",
+      "decorators": true
+    },
+    "target": "esnext",
+    "transform": {
+      "decoratorVersion": "2022-03"
+    }
+  },
+  "module": {
+    "type": "commonjs",
+    "ignoreDynamic": true
+  }
+}

.eslintrc.js

@@ -11,5 +11,13 @@ module.exports = {
     "max-params": ["error", 4],
     "prefer-const": "error",
   },
-  ignorePatterns: ["dist/"],
+  overrides: [
+    {
+      files: ["examples/**/*.ts"],
+      rules: {
+        "turbo/no-undeclared-env-vars": "off",
+      },
+    },
+  ],
+  ignorePatterns: ["dist/", "lib/"],
 };

.github/workflows/e2e.yml

@@ -1,68 +0,0 @@
-name: E2E Tests
-on:
-  push:
-    branches: [main]
-  pull_request:
-    paths:
-      - "packages/create-llama/**"
-      - ".github/workflows/e2e.yml"
-    branches: [main]
-
-env:
-  POETRY_VERSION: "1.6.1"
-
-jobs:
-  e2e:
-    name: create-llama
-    timeout-minutes: 60
-    strategy:
-      fail-fast: true
-      matrix:
-        node-version: [18, 20]
-        python-version: ["3.11"]
-        os: [macos-latest, windows-latest]
-    defaults:
-      run:
-        shell: bash
-    runs-on: ${{ matrix.os }}
-    steps:
-      - uses: actions/checkout@v4
-      - name: Set up python ${{ matrix.python-version }}
-        uses: actions/setup-python@v4
-        with:
-          python-version: ${{ matrix.python-version }}
-      - name: Install Poetry
-        uses: snok/install-poetry@v1
-        with:
-          version: ${{ env.POETRY_VERSION }}
-      - uses: pnpm/action-setup@v2
-      - name: Setup Node.js ${{ matrix.node-version }}
-        uses: actions/setup-node@v4
-        with:
-          node-version: ${{ matrix.node-version }}
-          cache: "pnpm"
-      - name: Install dependencies
-        run: pnpm install
-      - name: Install Playwright Browsers
-        run: pnpm exec playwright install --with-deps
-        working-directory: ./packages/create-llama
-      - name: Build create-llama
-        run: pnpm run build
-        working-directory: ./packages/create-llama
-      - name: Pack
-        run: pnpm pack --pack-destination ./output
-        working-directory: ./packages/create-llama
-      - name: Extract Pack
-        run: tar -xvzf ./output/*.tgz -C ./output
-        working-directory: ./packages/create-llama
-      - name: Run Playwright tests
-        run: pnpm exec playwright test
-        env:
-          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
-        working-directory: ./packages/create-llama
-      - uses: actions/upload-artifact@v3
-        if: always()
-        with:
-          name: playwright-report
-          path: ./packages/create-llama/playwright-report/
-          retention-days: 30

.github/workflows/publish.yml

@@ -0,0 +1,36 @@
+name: Publish
+on:
+  push:
+    branches:
+      - main
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+
+    permissions:
+      contents: read
+      id-token: write
+
+    steps:
+      - uses: actions/checkout@v4
+      - uses: pnpm/action-setup@v2
+      - name: Setup Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version-file: ".nvmrc"
+          cache: "pnpm"
+      - name: Install dependencies
+        run: pnpm install
+
+      - name: Publish @llamaindex/env
+        run: npx jsr publish
+        working-directory: packages/env
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Publish @llamaindex/core
+        run: npx jsr publish --allow-slow-types
+        working-directory: packages/core
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/test.yml

@@ -1,9 +1,44 @@
 name: Run Tests
 
-on: [push, pull_request]
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+    branches:
+      - main
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
 
 jobs:
+  e2e:
+    strategy:
+      fail-fast: false
+      matrix:
+        node-version: [18.x, 20.x, 21.x]
+    name: E2E on Node.js ${{ matrix.node-version }}
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: pnpm/action-setup@v2
+      - name: Setup Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version-file: ".nvmrc"
+          cache: "pnpm"
+      - name: Install dependencies
+        run: pnpm install
+      - name: Run E2E Tests
+        run: pnpm run e2e
+
   test:
+    strategy:
+      fail-fast: false
+      matrix:
+        node-version: [18.x, 20.x, 21.x]
+    name: Test on Node.js ${{ matrix.node-version }}
     runs-on: ubuntu-latest
 
     steps:
@@ -12,7 +47,7 @@ jobs:
       - name: Setup Node.js
         uses: actions/setup-node@v4
         with:
-          node-version-file: ".nvmrc"
+          node-version: ${{ matrix.node-version }}
           cache: "pnpm"
       - name: Install dependencies
         run: pnpm install
@@ -44,6 +79,24 @@ jobs:
           name: typecheck-build-dist
           path: ./packages/core/dist
           if-no-files-found: error
+  core-edge-runtime:
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+      - uses: pnpm/action-setup@v2
+      - name: Setup Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version-file: ".nvmrc"
+          cache: "pnpm"
+      - name: Install dependencies
+        run: pnpm install
+      - name: Build
+        run: pnpm run build --filter @llamaindex/edge
+      - name: Build Edge Runtime
+        run: pnpm run build
+        working-directory: ./packages/edge/e2e/test-edge-runtime
   typecheck-examples:
     runs-on: ubuntu-latest
 

.gitignore

@@ -44,6 +44,7 @@ test-results/
 playwright-report/
 blob-report/
 playwright/.cache/
+.tsbuildinfo
 
 # intellij
 **/.idea

.swcrc

@@ -0,0 +1,12 @@
+{
+  "jsc": {
+    "parser": {
+      "syntax": "typescript",
+      "decorators": true
+    },
+    "target": "esnext",
+    "transform": {
+      "decoratorVersion": "2022-03"
+    }
+  }
+}

CONTRIBUTING.md

@@ -79,13 +79,27 @@ That should start a webserver which will serve the docs on https://localhost:300
 
 Any changes you make should be reflected in the browser. If you need to regenerate the API docs and find that your TSDoc isn't getting the updates, feel free to remove apps/docs/api. It will automatically regenerate itself when you run pnpm start again.
 
-## Publishing
+## Changeset
 
-To publish a new version of the library, run
+We use [changesets](https://github.com/changesets/changesets) for managing versions and changelogs. To create a new changeset, run:
+
+```
+pnpm changeset
+```
+
+Please send a descriptive changeset for each PR.
+
+## Publishing (maintainers only)
+
+To publish a new version of the library, first create a new version:
+
+```shell
+pnpm new-version
+```
+
+If everything looks good, commit the generated files and release the new version:
 
 ```shell
-pnpm new-llamaindex
-pnpm new-create-llama
 pnpm release
 git push # push to the main branch
 git push --tags

README.md

@@ -83,30 +83,38 @@ Check out our NextJS playground at https://llama-playground.vercel.app/. The sou
 
 - [Node](/packages/core/src/Node.ts): The basic data building block. Most commonly, these are parts of the document split into manageable pieces that are small enough to be fed into an embedding model and LLM.
 
-- [Embedding](/packages/core/src/Embedding.ts): Embeddings are sets of floating point numbers which represent the data in a Node. By comparing the similarity of embeddings, we can derive an understanding of the similarity of two pieces of data. One use case is to compare the embedding of a question with the embeddings of our Nodes to see which Nodes may contain the data needed to answer that quesiton.
+- [Embedding](/packages/core/src/embeddings/OpenAIEmbedding.ts): Embeddings are sets of floating point numbers which represent the data in a Node. By comparing the similarity of embeddings, we can derive an understanding of the similarity of two pieces of data. One use case is to compare the embedding of a question with the embeddings of our Nodes to see which Nodes may contain the data needed to answer that quesiton. Because the default service context is OpenAI, the default embedding is `OpenAIEmbedding`. If using different models, say through Ollama, use this [Embedding](/packages/core/src/embeddings/OllamaEmbedding.ts) (see all [here](/packages/core/src/embeddings)).
 
 - [Indices](/packages/core/src/indices/): Indices store the Nodes and the embeddings of those nodes. QueryEngines retrieve Nodes from these Indices using embedding similarity.
 
-- [QueryEngine](/packages/core/src/QueryEngine.ts): Query engines are what generate the query you put in and give you back the result. Query engines generally combine a pre-built prompt with selected Nodes from your Index to give the LLM the context it needs to answer your query.
+- [QueryEngine](/packages/core/src/engines/query/RetrieverQueryEngine.ts): Query engines are what generate the query you put in and give you back the result. Query engines generally combine a pre-built prompt with selected Nodes from your Index to give the LLM the context it needs to answer your query. To build a query engine from your Index (recommended), use the [`asQueryEngine`](/packages/core/src/indices/BaseIndex.ts) method on your Index. See all query engines [here](/packages/core/src/engines/query).
 
-- [ChatEngine](/packages/core/src/ChatEngine.ts): A ChatEngine helps you build a chatbot that will interact with your Indices.
+- [ChatEngine](/packages/core/src/engines/chat/SimpleChatEngine.ts): A ChatEngine helps you build a chatbot that will interact with your Indices. See all chat engines [here](/packages/core/src/engines/chat).
 
 - [SimplePrompt](/packages/core/src/Prompt.ts): A simple standardized function call definition that takes in inputs and formats them in a template literal. SimplePrompts can be specialized using currying and combined using other SimplePrompt functions.
 
-## Note: NextJS:
+## Using NextJS
 
-If you're using NextJS App Router, you'll need to use the NodeJS runtime (default) and add the following config to your next.config.js to have it use imports/exports in the same way Node does.
+If you're using the NextJS App Router, you can choose between the Node.js and the [Edge runtime](https://nextjs.org/docs/app/building-your-application/rendering/edge-and-nodejs-runtimes#edge-runtime).
 
-```js
-export const runtime = "nodejs"; // default
+With NextJS 13 and 14, using the Node.js runtime is the default. You can explicitly set the Edge runtime in your [router handler](https://nextjs.org/docs/app/building-your-application/routing/route-handlers) by adding this line:
+
+```typescript
+export const runtime = "edge";
 ```
 
+The following sections explain further differences in using the Node.js or Edge runtime.
+
+### Using the Node.js runtime
+
+Add the following config to your `next.config.js` to ignore specific packages in the server-side bundling:
+
 ```js
 // next.config.js
 /** @type {import('next').NextConfig} */
 const nextConfig = {
   experimental: {
-    serverComponentsExternalPackages: ["pdf2json"],
+    serverComponentsExternalPackages: ["pdf2json", "@zilliz/milvus2-sdk-node"],
   },
   webpack: (config) => {
     config.resolve.alias = {
@@ -121,10 +129,60 @@ const nextConfig = {
 module.exports = nextConfig;
 ```
 
+### Using the Edge runtime
+
+We publish a dedicated package (`@llamaindex/edge` instead of `llamaindex`) for using the Edge runtime. To use it, first install the package:
+
+```shell
+pnpm install @llamaindex/edge
+```
+
+> _Note_: Ensure that your `package.json` doesn't include the `llamaindex` package if you're using `@llamaindex/edge`.
+
+Then make sure to use the correct import statement in your code:
+
+```typescript
+// replace 'llamaindex' with '@llamaindex/edge'
+import {} from "@llamaindex/edge";
+```
+
+A further difference is that the `@llamaindex/edge` package doesn't export classes from the `readers` or `storage` folders. The reason is that most of these classes are not compatible with the Edge runtime.
+
+If you need any of those classes, you have to import them instead directly. Here's an example for importing the `PineconeVectorStore` class:
+
+```typescript
+import { PineconeVectorStore } from "@llamaindex/edge/storage/vectorStore/PineconeVectorStore";
+```
+
+As the `PDFReader` is not with the Edge runtime, here's how to use the `SimpleDirectoryReader` with the `LlamaParseReader` to load PDFs:
+
+```typescript
+import { SimpleDirectoryReader } from "@llamaindex/edge/readers/SimpleDirectoryReader";
+import { LlamaParseReader } from "@llamaindex/edge/readers/LlamaParseReader";
+
+export const DATA_DIR = "./data";
+
+export async function getDocuments() {
+  const reader = new SimpleDirectoryReader();
+  // Load PDFs using LlamaParseReader
+  return await reader.loadData({
+    directoryPath: DATA_DIR,
+    fileExtToReader: {
+      pdf: new LlamaParseReader({ resultType: "markdown" }),
+    },
+  });
+}
+```
+
+> _Note_: Reader classes have to be added explictly to the `fileExtToReader` map in the Edge version of the `SimpleDirectoryReader`.
+
+You'll find a complete example of using the Edge runtime with LlamaIndexTS here: https://github.com/run-llama/create_llama_projects/tree/main/nextjs-edge-llamaparse
+
 ## Supported LLMs:
 
 - OpenAI GPT-3.5-turbo and GPT-4
-- Anthropic Claude Instant and Claude 2
+- Anthropic Claude 3 (Opus, Sonnet, and Haiku) and the legacy models (Claude 2 and Instant)
+- Groq LLMs
 - Llama2 Chat LLMs (70B, 13B, and 7B parameters)
 - MistralAI Chat LLMs
 - Fireworks Chat LLMs

apps/docs/CHANGELOG.md

@@ -1,5 +1,20 @@
 # docs
 
+## 0.0.4
+
+### Patch Changes
+
+- Updated dependencies [5116ad8]
+  - @llamaindex/env@0.0.5
+
+## 0.0.3
+
+### Patch Changes
+
+- 09bf27a: Add Groq LLM to LlamaIndex
+- Updated dependencies [cf87f84]
+  - @llamaindex/env@0.0.4
+
 ## 0.0.2
 
 ### Patch Changes

apps/docs/docs/getting_started/starter.md

@@ -1,56 +0,0 @@
----
-sidebar_position: 1
----
-
-# Starter Tutorial
-
-Once you have [installed LlamaIndex.TS using NPM](installation) and set up your OpenAI key, you're ready to start your first app:
-
-In a new folder:
-
-```bash npm2yarn
-npm install typescript
-npm install @types/node
-npx tsc --init # if needed
-```
-
-Create the file `example.ts`. This code will load some example data, create a document, index it (which creates embeddings using OpenAI), and then creates query engine to answer questions about the data.
-
-```ts
-// example.ts
-import fs from "fs/promises";
-import { Document, VectorStoreIndex } from "llamaindex";
-
-async function main() {
-  // Load essay from abramov.txt in Node
-  const essay = await fs.readFile(
-    "node_modules/llamaindex/examples/abramov.txt",
-    "utf-8",
-  );
-
-  // Create Document object with essay
-  const document = new Document({ text: essay });
-
-  // Split text and create embeddings. Store them in a VectorStoreIndex
-  const index = await VectorStoreIndex.fromDocuments([document]);
-
-  // Query the index
-  const queryEngine = index.asQueryEngine();
-  const response = await queryEngine.query({
-    query: "What did the author do in college?",
-  });
-
-  // Output response
-  console.log(response.toString());
-}
-
-main();
-```
-
-Then you can run it using
-
-```bash
-npx ts-node example.ts
-```
-
-Ready to learn more? Check out our NextJS playground at https://llama-playground.vercel.app/. The source is available at https://github.com/run-llama/ts-playground

apps/docs/docs/getting_started/starter.mdx

@@ -0,0 +1,51 @@
+---
+sidebar_position: 1
+---
+
+import CodeBlock from "@theme/CodeBlock";
+import CodeSource from "!raw-loader!../../../../examples/vectorIndex";
+import TSConfigSource from "!!raw-loader!../../../../examples/tsconfig.json";
+
+# Starter Tutorial
+
+Make sure you have installed LlamaIndex.TS and have an OpenAI key. If you haven't, check out the [installation](installation) guide.
+
+## From scratch(node.js + TypeScript):
+
+In a new folder:
+
+```bash npm2yarn
+npm init
+npm install -D typescript @types/node
+```
+
+Create the file `example.ts`. This code will load some example data, create a document, index it (which creates embeddings using OpenAI), and then creates query engine to answer questions about the data.
+
+<CodeBlock language="ts">{CodeSource}</CodeBlock>
+
+Create a `tsconfig.json` file in the same folder:
+
+<CodeBlock language="json">{TSConfigSource}</CodeBlock>
+
+Now you can run the code with
+
+```bash
+npx tsx example.ts
+```
+
+Also, you can clone our examples and try them out:
+
+```bash npm2yarn
+npx degit run-llama/LlamaIndexTS/examples my-new-project
+cd my-new-project
+npm install
+npx tsx ./vectorIndex.ts
+```
+
+## From scratch (Next.js + TypeScript):
+
+You just need one command to create a new Next.js project:
+
+```bash npm2yarn
+npx create-llama@latest
+```

apps/docs/docs/introduction.md

@@ -37,7 +37,7 @@ For more complex applications, our lower-level APIs allow advanced users to cust
 
 `npm install llamaindex`
 
-Our documentation includes [Installation Instructions](./getting_started/installation.mdx) and a [Starter Tutorial](./getting_started/starter.md) to build your first application.
+Our documentation includes [Installation Instructions](./getting_started/installation.mdx) and a [Starter Tutorial](./getting_started/starter.mdx) to build your first application.
 
 Once you're up and running, [High-Level Concepts](./getting_started/concepts.md) has an overview of LlamaIndex's modular architecture. For more hands-on practical examples, look through our Examples section on the sidebar.
 

apps/docs/docs/modules/agent/multi_document_agent.mdx

@@ -33,7 +33,7 @@ import {
   SimpleToolNodeMapping,
   SummaryIndex,
   VectorStoreIndex,
-  serviceContextFromDefaults,
+  Settings,
   storageContextFromDefaults,
 } from "llamaindex";
 ```
@@ -147,12 +147,10 @@ for (const title of wikiTitles) {
 We will be using gpt-4 for this example and we will use the `StorageContext` to store the documents in-memory.
 
 ```ts
-const llm = new OpenAI({
+Settings.llm = new OpenAI({
   model: "gpt-4",
 });
 
-const ctx = serviceContextFromDefaults({ llm });
-
 const storageContext = await storageContextFromDefaults({
   persistDir: "./storage",
 });
@@ -189,14 +187,12 @@ for (const title of wikiTitles) {
 
   // create the vector index for specific search
   const vectorIndex = await VectorStoreIndex.init({
-    serviceContext: serviceContext,
     storageContext: storageContext,
     nodes,
   });
 
   // create the summary index for broader search
   const summaryIndex = await SummaryIndex.init({
-    serviceContext: serviceContext,
     nodes,
   });
 
@@ -278,7 +274,6 @@ const objectIndex = await ObjectIndex.fromObjects(
   toolMapping,
   VectorStoreIndex,
   {
-    serviceContext,
     storageContext,
   },
 );

apps/docs/docs/modules/embeddings/available_embeddings/huggingface.md

@@ -3,17 +3,14 @@
 To use HuggingFace embeddings, you need to import `HuggingFaceEmbedding` from `llamaindex`.
 
 ```ts
-import { HuggingFaceEmbedding, serviceContextFromDefaults } from "llamaindex";
+import { HuggingFaceEmbedding, Settings } from "llamaindex";
 
-const huggingFaceEmbeds = new HuggingFaceEmbedding();
-
-const serviceContext = serviceContextFromDefaults({ embedModel: openaiEmbeds });
+// Update Embed Model
+Settings.embedModel = new HuggingFaceEmbedding();
 
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 
 const queryEngine = index.asQueryEngine();
 
@@ -23,3 +20,15 @@ const results = await queryEngine.query({
   query,
 });
 ```
+
+Per default, `HuggingFaceEmbedding` is using the `Xenova/all-MiniLM-L6-v2` model. You can change the model by passing the `modelType` parameter to the constructor.
+If you're not using a quantized model, set the `quantized` parameter to `false`.
+
+For example, to use the not quantized `BAAI/bge-small-en-v1.5` model, you can use the following code:
+
+```ts
+Settings.embedModel = new HuggingFaceEmbedding({
+  modelType: "BAAI/bge-small-en-v1.5",
+  quantized: false,
+});
+```

apps/docs/docs/modules/embeddings/available_embeddings/mistral.md

@@ -3,21 +3,16 @@
 To use MistralAI embeddings, you need to import `MistralAIEmbedding` from `llamaindex`.
 
 ```ts
-import { MistralAIEmbedding, serviceContextFromDefaults } from "llamaindex";
+import { MistralAIEmbedding, Settings } from "llamaindex";
 
-const mistralEmbedModel = new MistralAIEmbedding({
+// Update Embed Model
+Settings.embedModel = new MistralAIEmbedding({
   apiKey: "<YOUR_API_KEY>",
 });
 
-const serviceContext = serviceContextFromDefaults({
-  embedModel: mistralEmbedModel,
-});
-
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 
 const queryEngine = index.asQueryEngine();
 

apps/docs/docs/modules/embeddings/available_embeddings/ollama.md

@@ -3,19 +3,13 @@
 To use Ollama embeddings, you need to import `Ollama` from `llamaindex`.
 
 ```ts
-import { Ollama, serviceContextFromDefaults } from "llamaindex";
+import { Ollama, Settings } from "llamaindex";
 
-const ollamaEmbedModel = new Ollama();
-
-const serviceContext = serviceContextFromDefaults({
-  embedModel: ollamaEmbedModel,
-});
+Settings.embedModel = new Ollama();
 
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 
 const queryEngine = index.asQueryEngine();
 

apps/docs/docs/modules/embeddings/available_embeddings/openai.md

@@ -3,19 +3,13 @@
 To use OpenAI embeddings, you need to import `OpenAIEmbedding` from `llamaindex`.
 
 ```ts
-import { OpenAIEmbedding, serviceContextFromDefaults } from "llamaindex";
+import { OpenAIEmbedding, Settings } from "llamaindex";
 
-const openaiEmbedModel = new OpenAIEmbedding();
-
-const serviceContext = serviceContextFromDefaults({
-  embedModel: openaiEmbedModel,
-});
+Settings.embedModel = new OpenAIEmbedding();
 
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 
 const queryEngine = index.asQueryEngine();
 

apps/docs/docs/modules/embeddings/available_embeddings/together.md

@@ -3,21 +3,15 @@
 To use together embeddings, you need to import `TogetherEmbedding` from `llamaindex`.
 
 ```ts
-import { TogetherEmbedding, serviceContextFromDefaults } from "llamaindex";
+import { TogetherEmbedding, Settings } from "llamaindex";
 
-const togetherEmbedModel = new TogetherEmbedding({
+Settings.embedModel = new TogetherEmbedding({
   apiKey: "<YOUR_API_KEY>",
 });
 
-const serviceContext = serviceContextFromDefaults({
-  embedModel: togetherEmbedModel,
-});
-
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 
 const queryEngine = index.asQueryEngine();
 

apps/docs/docs/modules/embeddings/index.md

@@ -2,14 +2,14 @@
 
 The embedding model in LlamaIndex is responsible for creating numerical representations of text. By default, LlamaIndex will use the `text-embedding-ada-002` model from OpenAI.
 
-This can be explicitly set in the `ServiceContext` object.
+This can be explicitly updated through `Settings`
 
 ```typescript
-import { OpenAIEmbedding, serviceContextFromDefaults } from "llamaindex";
+import { OpenAIEmbedding, Settings } from "llamaindex";
 
-const openaiEmbeds = new OpenAIEmbedding();
-
-const serviceContext = serviceContextFromDefaults({ embedModel: openaiEmbeds });
+Settings.embedModel = new OpenAIEmbedding({
+  model: "text-embedding-ada-002",
+});
 ```
 
 ## Local Embedding
@@ -19,4 +19,3 @@ For local embeddings, you can use the [HuggingFace](./available_embeddings/huggi
 ## API Reference
 
 - [OpenAIEmbedding](../../api/classes/OpenAIEmbedding.md)
-- [ServiceContext](../../api/interfaces//ServiceContext.md)

apps/docs/docs/modules/evaluation/_category_.yml

@@ -0,0 +1,2 @@
+label: "Evaluating"
+position: 3

apps/docs/docs/modules/evaluation/index.md

@@ -0,0 +1,32 @@
+# Evaluating
+
+## Concept
+
+Evaluation and benchmarking are crucial concepts in LLM development. To improve the perfomance of an LLM app (RAG, agents) you must have a way to measure it.
+
+LlamaIndex offers key modules to measure the quality of generated results. We also offer key modules to measure retrieval quality.
+
+- **Response Evaluation**: Does the response match the retrieved context? Does it also match the query? Does it match the reference answer or guidelines?
+- **Retrieval Evaluation**: Are the retrieved sources relevant to the query?
+
+## Response Evaluation
+
+Evaluation of generated results can be difficult, since unlike traditional machine learning the predicted result is not a single number, and it can be hard to define quantitative metrics for this problem.
+
+LlamaIndex offers LLM-based evaluation modules to measure the quality of results. This uses a “gold” LLM (e.g. GPT-4) to decide whether the predicted answer is correct in a variety of ways.
+
+Note that many of these current evaluation modules do not require ground-truth labels. Evaluation can be done with some combination of the query, context, response, and combine these with LLM calls.
+
+These evaluation modules are in the following forms:
+
+- **Correctness**: Whether the generated answer matches that of the reference answer given the query (requires labels).
+
+- **Faithfulness**: Evaluates if the answer is faithful to the retrieved contexts (in other words, whether if there’s hallucination).
+
+- **Relevancy**: Evaluates if the response from a query engine matches any source nodes.
+
+## Usage
+
+- [Correctness Evaluator](./modules/correctness.md)
+- [Faithfulness Evaluator](./modules/faithfulness.md)
+- [Relevancy Evaluator](./modules/relevancy.md)

apps/docs/docs/modules/evaluation/modules/_category_.yml

@@ -0,0 +1 @@
+label: "Modules"

apps/docs/docs/modules/evaluation/modules/correctness.md

@@ -0,0 +1,58 @@
+# Correctness Evaluator
+
+Correctness evaluates the relevance and correctness of a generated answer against a reference answer.
+
+This is useful for measuring if the response was correct. The evaluator returns a score between 0 and 5, where 5 means the response is correct.
+
+## Usage
+
+Firstly, you need to install the package:
+
+```bash
+pnpm i llamaindex
+```
+
+Set the OpenAI API key:
+
+```bash
+export OPENAI_API_KEY=your-api-key
+```
+
+Import the required modules:
+
+```ts
+import { CorrectnessEvaluator, OpenAI, Settings } from "llamaindex";
+```
+
+Let's setup gpt-4 for better results:
+
+```ts
+Settings.llm = new OpenAI({
+  model: "gpt-4",
+});
+```
+
+```ts
+const query =
+  "Can you explain the theory of relativity proposed by Albert Einstein in detail?";
+
+const response = ` Certainly! Albert Einstein's theory of relativity consists of two main components: special relativity and general relativity. Special relativity, published in 1905, introduced the concept that the laws of physics are the same for all non-accelerating observers and that the speed of light in a vacuum is a constant, regardless of the motion of the source or observer. It also gave rise to the famous equation E=mc², which relates energy (E) and mass (m).
+
+However, general relativity, published in 1915, extended these ideas to include the effects of magnetism. According to general relativity, gravity is not a force between masses but rather the result of the warping of space and time by magnetic fields generated by massive objects. Massive objects, such as planets and stars, create magnetic fields that cause a curvature in spacetime, and smaller objects follow curved paths in response to this magnetic curvature. This concept is often illustrated using the analogy of a heavy ball placed on a rubber sheet with magnets underneath, causing it to create a depression that other objects (representing smaller masses) naturally move towards due to magnetic attraction.
+`;
+
+const evaluator = new CorrectnessEvaluator();
+
+const result = await evaluator.evaluateResponse({
+  query,
+  response,
+});
+
+console.log(
+  `the response is ${result.passing ? "correct" : "not correct"} with a score of ${result.score}`,
+);
+```
+
+```bash
+the response is not correct with a score of 2.5
+```

apps/docs/docs/modules/evaluation/modules/faithfulness.md

@@ -0,0 +1,78 @@
+# Faithfulness Evaluator
+
+Faithfulness is a measure of whether the generated answer is faithful to the retrieved contexts. In other words, it measures whether there is any hallucination in the generated answer.
+
+This uses the FaithfulnessEvaluator module to measure if the response from a query engine matches any source nodes.
+
+This is useful for measuring if the response was hallucinated. The evaluator returns a score between 0 and 1, where 1 means the response is faithful to the retrieved contexts.
+
+## Usage
+
+Firstly, you need to install the package:
+
+```bash
+pnpm i llamaindex
+```
+
+Set the OpenAI API key:
+
+```bash
+export OPENAI_API_KEY=your-api-key
+```
+
+Import the required modules:
+
+```ts
+import {
+  Document,
+  FaithfulnessEvaluator,
+  OpenAI,
+  VectorStoreIndex,
+  Settings,
+} from "llamaindex";
+```
+
+Let's setup gpt-4 for better results:
+
+```ts
+Settings.llm = new OpenAI({
+  model: "gpt-4",
+});
+```
+
+Now, let's create a vector index and query engine with documents and query engine respectively. Then, we can evaluate the response with the query and response from the query engine.:
+
+```ts
+const documents = [
+  new Document({
+    text: `The city came under British control in 1664 and was renamed New York after King Charles II of England granted the lands to his brother, the Duke of York. The city was regained by the Dutch in July 1673 and was renamed New Orange for one year and three months; the city has been continuously named New York since November 1674. New York City was the capital of the United States from 1785 until 1790, and has been the largest U.S. city since 1790. The Statue of Liberty greeted millions of immigrants as they came to the U.S. by ship in the late 19th and early 20th centuries, and is a symbol of the U.S. and its ideals of liberty and peace. In the 21st century, New York City has emerged as a global node of creativity, entrepreneurship, and as a symbol of freedom and cultural diversity. The New York Times has won the most Pulitzer Prizes for journalism and remains the U.S. media's "newspaper of record". In 2019, New York City was voted the greatest city in the world in a survey of over 30,000 p...	Pass`,
+  }),
+];
+
+const vectorIndex = await VectorStoreIndex.fromDocuments(documents);
+
+const queryEngine = vectorIndex.asQueryEngine();
+```
+
+Now, let's evaluate the response:
+
+```ts
+const query = "How did New York City get its name?";
+
+const evaluator = new FaithfulnessEvaluator();
+
+const response = await queryEngine.query({
+  query,
+});
+
+const result = await evaluator.evaluateResponse({
+  query,
+  response,
+});
+
+console.log(`the response is ${result.passing ? "faithful" : "not faithful"}`);
+```
+
+```bash
+the response is faithful
+```

apps/docs/docs/modules/evaluation/modules/relevancy.md

@@ -0,0 +1,66 @@
+# Relevancy Evaluator
+
+Relevancy measure if the response from a query engine matches any source nodes.
+
+It is useful for measuring if the response was relevant to the query. The evaluator returns a score between 0 and 1, where 1 means the response is relevant to the query.
+
+## Usage
+
+Firstly, you need to install the package:
+
+```bash
+pnpm i llamaindex
+```
+
+Set the OpenAI API key:
+
+```bash
+export OPENAI_API_KEY=your-api-key
+```
+
+Import the required modules:
+
+```ts
+import { RelevancyEvaluator, OpenAI, Settings } from "llamaindex";
+```
+
+Let's setup gpt-4 for better results:
+
+```ts
+Settings.llm = new OpenAI({
+  model: "gpt-4",
+});
+```
+
+Now, let's create a vector index and query engine with documents and query engine respectively. Then, we can evaluate the response with the query and response from the query engine.:
+
+```ts
+const documents = [
+  new Document({
+    text: `The city came under British control in 1664 and was renamed New York after King Charles II of England granted the lands to his brother, the Duke of York. The city was regained by the Dutch in July 1673 and was renamed New Orange for one year and three months; the city has been continuously named New York since November 1674. New York City was the capital of the United States from 1785 until 1790, and has been the largest U.S. city since 1790. The Statue of Liberty greeted millions of immigrants as they came to the U.S. by ship in the late 19th and early 20th centuries, and is a symbol of the U.S. and its ideals of liberty and peace. In the 21st century, New York City has emerged as a global node of creativity, entrepreneurship, and as a symbol of freedom and cultural diversity. The New York Times has won the most Pulitzer Prizes for journalism and remains the U.S. media's "newspaper of record". In 2019, New York City was voted the greatest city in the world in a survey of over 30,000 p...	Pass`,
+  }),
+];
+
+const vectorIndex = await VectorStoreIndex.fromDocuments(documents);
+
+const queryEngine = vectorIndex.asQueryEngine();
+
+const query = "How did New York City get its name?";
+
+const response = await queryEngine.query({
+  query,
+});
+
+const evaluator = new RelevancyEvaluator();
+
+const result = await evaluator.evaluateResponse({
+  query,
+  response: response,
+});
+
+console.log(`the response is ${result.passing ? "relevant" : "not relevant"}`);
+```
+
+```bash
+the response is relevant
+```

apps/docs/docs/modules/ingestion_pipeline/transformations.md

@@ -1,6 +1,6 @@
 # Transformations
 
-A transformation is something that takes a list of nodes as an input, and returns a list of nodes. Each component that implements the Transformatio class has both a `transform` definition responsible for transforming the nodes
+A transformation is something that takes a list of nodes as an input, and returns a list of nodes. Each component that implements the Transformation class has both a `transform` definition responsible for transforming the nodes.
 
 Currently, the following components are Transformation objects:
 

apps/docs/docs/modules/llms/available_llms/anthropic.md

@@ -3,13 +3,11 @@
 ## Usage
 
 ```ts
-import { Anthropic, serviceContextFromDefaults } from "llamaindex";
+import { Anthropic, Settings } from "llamaindex";
 
-const anthropicLLM = new Anthropic({
+Settings.llm = new Anthropic({
   apiKey: "<YOUR_API_KEY>",
 });
-
-const serviceContext = serviceContextFromDefaults({ llm: anthropicLLM });
 ```
 
 ## Load and index documents
@@ -19,9 +17,7 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Query
@@ -39,28 +35,17 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import {
-  Anthropic,
-  Document,
-  VectorStoreIndex,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { Anthropic, Document, VectorStoreIndex, Settings } from "llamaindex";
+
+Settings.llm = new Anthropic({
+  apiKey: "<YOUR_API_KEY>",
+});
 
 async function main() {
-  // Create an instance of the Anthropic LLM
-  const anthropicLLM = new Anthropic({
-    apiKey: "<YOUR_API_KEY>",
-  });
-
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({ llm: anthropicLLM });
-
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document], {
-    serviceContext,
-  });
+  const index = await VectorStoreIndex.fromDocuments([document]);
 
   // Create a query engine
   const queryEngine = index.asQueryEngine({

apps/docs/docs/modules/llms/available_llms/azure.md

@@ -15,11 +15,9 @@ export AZURE_OPENAI_DEPLOYMENT="gpt-4" # or some other deployment name
 ## Usage
 
 ```ts
-import { OpenAI, serviceContextFromDefaults } from "llamaindex";
+import { OpenAI, Settings } from "llamaindex";
 
-const azureOpenaiLLM = new OpenAI({ model: "gpt-4", temperature: 0 });
-
-const serviceContext = serviceContextFromDefaults({ llm: azureOpenaiLLM });
+Settings.llm = new OpenAI({ model: "gpt-4", temperature: 0 });
 ```
 
 ## Load and index documents
@@ -29,9 +27,7 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Query
@@ -49,26 +45,15 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import {
-  OpenAI,
-  Document,
-  VectorStoreIndex,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { OpenAI, Document, VectorStoreIndex, Settings } from "llamaindex";
+
+Settings.llm = new OpenAI({ model: "gpt-4", temperature: 0 });
 
 async function main() {
-  // Create an instance of the LLM
-  const azureOpenaiLLM = new OpenAI({ model: "gpt-4", temperature: 0 });
-
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({ llm: azureOpenaiLLM });
-
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document], {
-    serviceContext,
-  });
+  const index = await VectorStoreIndex.fromDocuments([document]);
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/fireworks.md

@@ -5,13 +5,11 @@ Fireworks.ai focus on production use cases for open source LLMs, offering speed
 ## Usage
 
 ```ts
-import { FireworksLLM, serviceContextFromDefaults } from "llamaindex";
+import { FireworksLLM, Settings } from "llamaindex";
 
-const fireworksLLM = new FireworksLLM({
+Settings.llm = new FireworksLLM({
   apiKey: "<YOUR_API_KEY>",
 });
-
-const serviceContext = serviceContextFromDefaults({ llm: fireworksLLM });
 ```
 
 ## Load and index documents
@@ -23,9 +21,7 @@ const reader = new PDFReader();
 const documents = await reader.loadData("../data/brk-2022.pdf");
 
 // Split text and create embeddings. Store them in a VectorStoreIndex
-const index = await VectorStoreIndex.fromDocuments(documents, {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments(documents);
 ```
 
 ## Query

apps/docs/docs/modules/llms/available_llms/groq.mdx

@@ -0,0 +1,52 @@
+import CodeBlock from "@theme/CodeBlock";
+import CodeSource from "!raw-loader!../../../../../../examples/groq.ts";
+
+# Groq
+
+## Usage
+
+First, create an API key at the [Groq Console](https://console.groq.com/keys). Then save it in your environment:
+
+```bash
+export GROQ_API_KEY=<your-api-key>
+```
+
+The initialize the Groq module.
+
+```ts
+import { Groq, Settings } from "llamaindex";
+
+Settings.llm = new Groq({
+  // If you do not wish to set your API key in the environment, you may
+  // configure your API key when you initialize the Groq class.
+  // apiKey: "<your-api-key>",
+});
+```
+
+## Load and index documents
+
+For this example, we will use a single document. In a real-world scenario, you would have multiple documents to index.
+
+```ts
+const document = new Document({ text: essay, id_: "essay" });
+
+const index = await VectorStoreIndex.fromDocuments([document]);
+```
+
+## Query
+
+```ts
+const queryEngine = index.asQueryEngine();
+
+const query = "What is the meaning of life?";
+
+const results = await queryEngine.query({
+  query,
+});
+```
+
+## Full Example
+
+<CodeBlock language="ts" showLineNumbers>
+  {CodeSource}
+</CodeBlock>

apps/docs/docs/modules/llms/available_llms/llama2.md

@@ -3,32 +3,24 @@
 ## Usage
 
 ```ts
-import { Ollama, serviceContextFromDefaults } from "llamaindex";
+import { Ollama, Settings } from "llamaindex";
 
-const llama2LLM = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
-
-const serviceContext = serviceContextFromDefaults({ llm: llama2LLM });
+Settings.llm = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
 ```
 
 ## Usage with Replication
 
 ```ts
-import {
-  Ollama,
-  ReplicateSession,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { Ollama, ReplicateSession, Settings } from "llamaindex";
 
 const replicateSession = new ReplicateSession({
   replicateKey,
 });
 
-const llama2LLM = new LlamaDeuce({
+Settings.llm = new LlamaDeuce({
   chatStrategy: DeuceChatStrategy.META,
   replicateSession,
 });
-
-const serviceContext = serviceContextFromDefaults({ llm: llama2LLM });
 ```
 
 ## Load and index documents
@@ -38,9 +30,7 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Query
@@ -58,26 +48,16 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import {
-  LlamaDeuce,
-  Document,
-  VectorStoreIndex,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { LlamaDeuce, Document, VectorStoreIndex, Settings } from "llamaindex";
+
+// Use the LlamaDeuce LLM
+Settings.llm = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
 
 async function main() {
-  // Create an instance of the LLM
-  const llama2LLM = new LlamaDeuce({ chatStrategy: DeuceChatStrategy.META });
-
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({ llm: mistralLLM });
-
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document], {
-    serviceContext,
-  });
+  const index = await VectorStoreIndex.fromDocuments([document]);
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/mistral.md

@@ -3,14 +3,12 @@
 ## Usage
 
 ```ts
-import { Ollama, serviceContextFromDefaults } from "llamaindex";
+import { Ollama, Settings } from "llamaindex";
 
-const mistralLLM = new MistralAI({
+Settings.llm = new MistralAI({
   model: "mistral-tiny",
   apiKey: "<YOUR_API_KEY>",
 });
-
-const serviceContext = serviceContextFromDefaults({ llm: mistralLLM });
 ```
 
 ## Load and index documents
@@ -20,9 +18,7 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Query
@@ -40,26 +36,16 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import {
-  MistralAI,
-  Document,
-  VectorStoreIndex,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { MistralAI, Document, VectorStoreIndex, Settings } from "llamaindex";
+
+// Use the MistralAI LLM
+Settings.llm = new MistralAI({ model: "mistral-tiny" });
 
 async function main() {
-  // Create an instance of the LLM
-  const mistralLLM = new MistralAI({ model: "mistral-tiny" });
-
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({ llm: mistralLLM });
-
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document], {
-    serviceContext,
-  });
+  const index = await VectorStoreIndex.fromDocuments([document]);
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/ollama.md

@@ -3,14 +3,10 @@
 ## Usage
 
 ```ts
-import { Ollama, serviceContextFromDefaults } from "llamaindex";
+import { Ollama, Settings } from "llamaindex";
 
-const ollamaLLM = new Ollama({ model: "llama2", temperature: 0.75 });
-
-const serviceContext = serviceContextFromDefaults({
-  llm: ollamaLLM,
-  embedModel: ollamaLLM,
-});
+Settings.llm = ollamaLLM;
+Settings.embedModel = ollamaLLM;
 ```
 
 ## Load and index documents
@@ -20,9 +16,7 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Query
@@ -40,33 +34,23 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import {
-  Ollama,
-  Document,
-  VectorStoreIndex,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { Ollama, Document, VectorStoreIndex, Settings } from "llamaindex";
 
 import fs from "fs/promises";
 
+const ollama = new Ollama({ model: "llama2", temperature: 0.75 });
+
+// Use Ollama LLM and Embed Model
+Settings.llm = ollama;
+Settings.embedModel = ollama;
+
 async function main() {
-  // Create an instance of the LLM
-  const ollamaLLM = new Ollama({ model: "llama2", temperature: 0.75 });
-
   const essay = await fs.readFile("./paul_graham_essay.txt", "utf-8");
 
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({
-    embedModel: ollamaLLM, // prevent 'Set OpenAI Key in OPENAI_API_KEY env variable' error
-    llm: ollamaLLM,
-  });
-
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document], {
-    serviceContext,
-  });
+  const index = await VectorStoreIndex.fromDocuments([document]);
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/openai.md

@@ -1,11 +1,9 @@
 # OpenAI
 
 ```ts
-import { OpenAI, serviceContextFromDefaults } from "llamaindex";
+import { OpenAI, Settings } from "llamaindex";
 
-const openaiLLM = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0, apiKey: <YOUR_API_KEY> });
-
-const serviceContext = serviceContextFromDefaults({ llm: openaiLLM });
+Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0, apiKey: <YOUR_API_KEY> });
 ```
 
 You can setup the apiKey on the environment variables, like:
@@ -21,9 +19,7 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Query
@@ -41,26 +37,16 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import {
-  OpenAI,
-  Document,
-  VectorStoreIndex,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { OpenAI, Document, VectorStoreIndex, Settings } from "llamaindex";
+
+// Use the OpenAI LLM
+Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0 });
 
 async function main() {
-  // Create an instance of the LLM
-  const openaiLLM = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0 });
-
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({ llm: openaiLLM });
-
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document], {
-    serviceContext,
-  });
+  const index = await VectorStoreIndex.fromDocuments([document]);
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/portkey.md

@@ -3,13 +3,11 @@
 ## Usage
 
 ```ts
-import { Portkey, serviceContextFromDefaults } from "llamaindex";
+import { Portkey, Settings } from "llamaindex";
 
-const portkeyLLM = new Portkey({
+Settings.llm = new Portkey({
   apiKey: "<YOUR_API_KEY>",
 });
-
-const serviceContext = serviceContextFromDefaults({ llm: portkeyLLM });
 ```
 
 ## Load and index documents
@@ -19,9 +17,7 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Query
@@ -39,28 +35,19 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import {
-  Portkey,
-  Document,
-  VectorStoreIndex,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { Portkey, Document, VectorStoreIndex, Settings } from "llamaindex";
+
+// Use the Portkey LLM
+Settings.llm = new Portkey({
+  apiKey: "<YOUR_API_KEY>",
+});
 
 async function main() {
-  // Create an instance of the LLM
-  const portkeyLLM = new Portkey({
-    apiKey: "<YOUR_API_KEY>",
-  });
-
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({ llm: portkeyLLM });
-
+  // Create a document
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document], {
-    serviceContext,
-  });
+  const index = await VectorStoreIndex.fromDocuments([document]);
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/available_llms/together.md

@@ -3,13 +3,11 @@
 ## Usage
 
 ```ts
-import { TogetherLLM, serviceContextFromDefaults } from "llamaindex";
+import { TogetherLLM, Settings } from "llamaindex";
 
-const togetherLLM = new TogetherLLM({
+Settings.llm = new TogetherLLM({
   apiKey: "<YOUR_API_KEY>",
 });
-
-const serviceContext = serviceContextFromDefaults({ llm: togetherLLM });
 ```
 
 ## Load and index documents
@@ -19,9 +17,7 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Query
@@ -39,28 +35,17 @@ const results = await queryEngine.query({
 ## Full Example
 
 ```ts
-import {
-  TogetherLLM,
-  Document,
-  VectorStoreIndex,
-  serviceContextFromDefaults,
-} from "llamaindex";
+import { TogetherLLM, Document, VectorStoreIndex, Settings } from "llamaindex";
+
+Settings.llm = new TogetherLLM({
+  apiKey: "<YOUR_API_KEY>",
+});
 
 async function main() {
-  // Create an instance of the LLM
-  const togetherLLM = new TogetherLLM({
-    apiKey: "<YOUR_API_KEY>",
-  });
-
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({ llm: togetherLLM });
-
   const document = new Document({ text: essay, id_: "essay" });
 
   // Load and index documents
-  const index = await VectorStoreIndex.fromDocuments([document], {
-    serviceContext,
-  });
+  const index = await VectorStoreIndex.fromDocuments([document]);
 
   // get retriever
   const retriever = index.asRetriever();

apps/docs/docs/modules/llms/index.md

@@ -6,14 +6,12 @@ sidebar_position: 3
 
 The LLM is responsible for reading text and generating natural language responses to queries. By default, LlamaIndex.TS uses `gpt-3.5-turbo`.
 
-The LLM can be explicitly set in the `ServiceContext` object.
+The LLM can be explicitly updated through `Settings`.
 
 ```typescript
-import { OpenAI, serviceContextFromDefaults } from "llamaindex";
+import { OpenAI, Settings } from "llamaindex";
 
-const openaiLLM = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0 });
-
-const serviceContext = serviceContextFromDefaults({ llm: openaiLLM });
+Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0 });
 ```
 
 ## Azure OpenAI
@@ -35,4 +33,3 @@ For local LLMs, currently we recommend the use of [Ollama](./available_llms/olla
 ## API Reference
 
 - [OpenAI](../api/classes/OpenAI.md)
-- [ServiceContext](../api/interfaces//ServiceContext.md)

apps/docs/docs/modules/node_parser.md

@@ -4,15 +4,14 @@ sidebar_position: 4
 
 # NodeParser
 
-The `NodeParser` in LlamaIndex is responsible for splitting `Document` objects into more manageable `Node` objects. When you call `.fromDocuments()`, the `NodeParser` from the `ServiceContext` is used to do this automatically for you. Alternatively, you can use it to split documents ahead of time.
+The `NodeParser` in LlamaIndex is responsible for splitting `Document` objects into more manageable `Node` objects. When you call `.fromDocuments()`, the `NodeParser` from the `Settings` is used to do this automatically for you. Alternatively, you can use it to split documents ahead of time.
 
 ```typescript
 import { Document, SimpleNodeParser } from "llamaindex";
 
 const nodeParser = new SimpleNodeParser();
-const nodes = nodeParser.getNodesFromDocuments([
-  new Document({ text: "I am 10 years old. John is 20 years old." }),
-]);
+
+Settings.nodeParser = nodeParser;
 ```
 
 ## TextSplitter

apps/docs/docs/modules/node_postprocessors/cohere_reranker.md

@@ -18,7 +18,7 @@ import {
   Document,
   OpenAI,
   VectorStoreIndex,
-  serviceContextFromDefaults,
+  Settings,
 } from "llamaindex";
 ```
 
@@ -29,13 +29,9 @@ For this example, we will use a single document. In a real-world scenario, you w
 ```ts
 const document = new Document({ text: essay, id_: "essay" });
 
-const serviceContext = serviceContextFromDefaults({
-  llm: new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 }),
-});
+Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 });
 
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 ```
 
 ## Increase similarity topK to retrieve more results

apps/docs/docs/modules/node_postprocessors/index.md

@@ -36,7 +36,7 @@ const processor = new SimilarityPostprocessor({
   similarityCutoff: 0.7,
 });
 
-const filteredNodes = processor.postprocessNodes(nodes);
+const filteredNodes = await processor.postprocessNodes(nodes);
 
 // cohere rerank: rerank nodes given query using trained model
 const reranker = new CohereRerank({
@@ -58,7 +58,10 @@ Most commonly, node-postprocessors will be used in a query engine, where they ar
 ### Using Node Postprocessors in a Query Engine
 
 ```ts
-import { Node, NodeWithScore, SimilarityPostprocessor, CohereRerank } from "llamaindex";
+import { Node, NodeWithScore, SimilarityPostprocessor, CohereRerank, Settings } from "llamaindex";
+
+// Use OpenAI LLM
+Settings.llm = new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 });
 
 const nodes: NodeWithScore[] = [
   {
@@ -79,14 +82,6 @@ const reranker = new CohereRerank({
 
 const document = new Document({ text: "essay", id_: "essay" });
 
-const serviceContext = serviceContextFromDefaults({
-  llm: new OpenAI({ model: "gpt-3.5-turbo", temperature: 0.1 }),
-});
-
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
-
 const queryEngine = index.asQueryEngine({
   nodePostprocessors: [processor, reranker],
 });
@@ -100,7 +95,7 @@ const response = await queryEngine.query("<user_query>");
 ```ts
 import { SimilarityPostprocessor } from "llamaindex";
 
-nodes = await index.asRetriever().retrieve("test query str");
+nodes = await index.asRetriever().retrieve({ query: "test query str" });
 
 const processor = new SimilarityPostprocessor({
   similarityCutoff: 0.7,

apps/docs/docs/modules/prompt/index.md

@@ -31,13 +31,11 @@ The first method is to create a new instance of `ResponseSynthesizer` (or the mo
 ```ts
 // Create an instance of response synthesizer
 const responseSynthesizer = new ResponseSynthesizer({
-  responseBuilder: new CompactAndRefine(serviceContext, newTextQaPrompt),
+  responseBuilder: new CompactAndRefine(undefined, newTextQaPrompt),
 });
 
 // Create index
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 
 // Query the index
 const queryEngine = index.asQueryEngine({ responseSynthesizer });
@@ -53,9 +51,7 @@ The second method is that most of the modules in LlamaIndex have a `getPrompts`
 
 ```ts
 // Create index
-const index = await VectorStoreIndex.fromDocuments([document], {
-  serviceContext,
-});
+const index = await VectorStoreIndex.fromDocuments([document]);
 
 // Query the index
 const queryEngine = index.asQueryEngine();

apps/docs/docs/modules/query_engines/metadata_filtering.md

@@ -54,12 +54,13 @@ You can create a `ChromaVectorStore` to store the documents:
 
 ```ts
 const chromaVS = new ChromaVectorStore({ collectionName });
-const serviceContext = await storageContextFromDefaults({
+
+const storageContext = await storageContextFromDefaults({
   vectorStore: chromaVS,
 });
 
 const index = await VectorStoreIndex.fromDocuments(docs, {
-  storageContext: serviceContext,
+  storageContext: storageContext,
 });
 ```
 

apps/docs/docs/modules/query_engines/router_query_engine.md

@@ -18,7 +18,7 @@ import {
   SimpleNodeParser,
   SummaryIndex,
   VectorStoreIndex,
-  serviceContextFromDefaults,
+  Settings,
 } from "llamaindex";
 ```
 
@@ -34,17 +34,13 @@ const documents = await new SimpleDirectoryReader().loadData({
 
 ## Service Context
 
-Next, we need to define some basic rules and parse the documents into nodes. We will use the `SimpleNodeParser` to parse the documents into nodes and `ServiceContext` to define the rules (eg. LLM API key, chunk size, etc.):
+Next, we need to define some basic rules and parse the documents into nodes. We will use the `SimpleNodeParser` to parse the documents into nodes and `Settings` to define the rules (eg. LLM API key, chunk size, etc.):
 
 ```ts
-const nodeParser = new SimpleNodeParser({
+Settings.llm = new OpenAI();
+Settings.nodeParser = new SimpleNodeParser({
   chunkSize: 1024,
 });
-
-const serviceContext = serviceContextFromDefaults({
-  nodeParser,
-  llm: new OpenAI(),
-});
 ```
 
 ## Creating Indices
@@ -52,13 +48,8 @@ const serviceContext = serviceContextFromDefaults({
 Next, we need to create some indices. We will create a `VectorStoreIndex` and a `SummaryIndex`:
 
 ```ts
-const vectorIndex = await VectorStoreIndex.fromDocuments(documents, {
-  serviceContext,
-});
-
-const summaryIndex = await SummaryIndex.fromDocuments(documents, {
-  serviceContext,
-});
+const vectorIndex = await VectorStoreIndex.fromDocuments(documents);
+const summaryIndex = await SummaryIndex.fromDocuments(documents);
 ```
 
 ## Creating Query Engines
@@ -88,7 +79,6 @@ const queryEngine = RouterQueryEngine.fromDefaults({
       description: "Useful for retrieving specific context from Abramov",
     },
   ],
-  serviceContext,
 });
 ```
 
@@ -117,34 +107,23 @@ import {
   SimpleNodeParser,
   SummaryIndex,
   VectorStoreIndex,
-  serviceContextFromDefaults,
+  Settings,
 } from "llamaindex";
 
+Settings.llm = new OpenAI();
+Settings.nodeParser = new SimpleNodeParser({
+  chunkSize: 1024,
+});
+
 async function main() {
   // Load documents from a directory
   const documents = await new SimpleDirectoryReader().loadData({
     directoryPath: "node_modules/llamaindex/examples",
   });
 
-  // Parse the documents into nodes
-  const nodeParser = new SimpleNodeParser({
-    chunkSize: 1024,
-  });
-
-  // Create a service context
-  const serviceContext = serviceContextFromDefaults({
-    nodeParser,
-    llm: new OpenAI(),
-  });
-
   // Create indices
-  const vectorIndex = await VectorStoreIndex.fromDocuments(documents, {
-    serviceContext,
-  });
-
-  const summaryIndex = await SummaryIndex.fromDocuments(documents, {
-    serviceContext,
-  });
+  const vectorIndex = await VectorStoreIndex.fromDocuments(documents);
+  const summaryIndex = await SummaryIndex.fromDocuments(documents);
 
   // Create query engines
   const vectorQueryEngine = vectorIndex.asQueryEngine();
@@ -162,7 +141,6 @@ async function main() {
         description: "Useful for retrieving specific context from Abramov",
       },
     ],
-    serviceContext,
   });
 
   // Query the router query engine

apps/docs/docs/modules/retriever.md

@@ -11,7 +11,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Fetch nodes!
-const nodesWithScore = await retriever.retrieve("query string");
+const nodesWithScore = await retriever.retrieve({ query: "query string" });
 ```
 
 ## API Reference

apps/docs/docs/recipes/_category_.yml

@@ -0,0 +1,2 @@
+label: Recipes
+position: 3

apps/docs/docs/recipes/cost-analysis.mdx

@@ -0,0 +1,14 @@
+# Cost Analysis
+
+This page shows how to track LLM cost using APIs.
+
+## Callback Manager
+
+The callback manager is a class that manages the callback functions.
+
+You can register `llm-start`, `llm-end`, and `llm-stream` callbacks to the callback manager for tracking the cost.
+
+import CodeBlock from "@theme/CodeBlock";
+import CodeSource from "!raw-loader!../../../../examples/recipes/cost-analysis";
+
+<CodeBlock language="ts">{CodeSource}</CodeBlock>

apps/docs/i18n/ar/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // جلب العقد!
-const nodesWithScore = await retriever.retrieve("سلسلة الاستعلام");
+const nodesWithScore = await retriever.retrieve({ query: "سلسلة الاستعلام" });
 ```
 
 ## مرجع الواجهة البرمجية (API Reference)

apps/docs/i18n/bg/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Извличане на върхове!
-const nodesWithScore = await retriever.retrieve("query string");
+const nodesWithScore = await retriever.retrieve({ query: "query string" });
 ```
 
 ## API Reference (API справка)

apps/docs/i18n/cat/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const recuperador = vector_index.asRetriever();
 recuperador.similarityTopK = 3;
 
 // Obteniu els nodes!
-const nodesAmbPuntuació = await recuperador.retrieve("cadena de consulta");
+const nodesAmbPuntuació = await recuperador.retrieve({ query: "cadena de consulta" });
 ```
 
 ## Referència de l'API

apps/docs/i18n/cs/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Získání uzlů!
-const nodesWithScore = await retriever.retrieve("dotazovací řetězec");
+const nodesWithScore = await retriever.retrieve({ query: "dotazovací řetězec" });
 ```
 
 ## API Reference (Odkazy na rozhraní)

apps/docs/i18n/da/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Hent noder!
-const nodesWithScore = await retriever.retrieve("forespørgselsstreng");
+const nodesWithScore = await retriever.retrieve({ query: "forespørgselsstreng" });
 ```
 
 ## API Reference

apps/docs/i18n/de/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Knoten abrufen!
-const nodesWithScore = await retriever.retrieve("Abfragezeichenfolge");
+const nodesWithScore = await retriever.retrieve({ query: "Abfragezeichenfolge" });
 ```
 
 ## API-Referenz

apps/docs/i18n/el/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Ανάκτηση κόμβων!
-const nodesWithScore = await retriever.retrieve("συμβολοσειρά ερωτήματος");
+const nodesWithScore = await retriever.retrieve({ query: "συμβολοσειρά ερωτήματος" });
 ```
 
 ## Αναφορά API

apps/docs/i18n/es/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const recuperador = vector_index.asRetriever();
 recuperador.similarityTopK = 3;
 
 // ¡Obtener nodos!
-const nodosConPuntuación = await recuperador.retrieve("cadena de consulta");
+const nodosConPuntuación = await recuperador.retrieve({ query: "cadena de consulta" });
 ```
 
 ## Referencia de la API

apps/docs/i18n/et/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Too sõlmed!
-const nodesWithScore = await retriever.retrieve("päringu string");
+const nodesWithScore = await retriever.retrieve({ query: "päringu string" });
 ```
 
 ## API viide

apps/docs/i18n/fa/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // بازیابی گره ها!
-const nodesWithScore = await retriever.retrieve("رشته پرس و جو");
+const nodesWithScore = await retriever.retrieve({ query: "رشته پرس و جو" });
 ```
 
 ## مرجع API

apps/docs/i18n/fi/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Hae solmut!
-const nodesWithScore = await retriever.retrieve("kyselymerkkijono");
+const nodesWithScore = await retriever.retrieve({ query: "kyselymerkkijono" });
 ```
 
 ## API-viite

apps/docs/i18n/fr/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -11,7 +11,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Récupérer les nœuds !
-const nodesWithScore = await retriever.retrieve("chaîne de requête");
+const nodesWithScore = await retriever.retrieve({ query: "chaîne de requête" });
 ```
 
 ## Référence de l'API

apps/docs/i18n/he/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // אחזור צמתים!
-const nodesWithScore = await retriever.retrieve("מחרוזת שאילתה");
+const nodesWithScore = await retriever.retrieve({ query: "מחרוזת שאילתה" });
 ```
 
 ## מדריך לממשק API

apps/docs/i18n/hi/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // नोड्स प्राप्त करें!
-const nodesWithScore = await retriever.retrieve("क्वेरी स्ट्रिंग");
+const nodesWithScore = await retriever.retrieve({ query: "क्वेरी स्ट्रिंग" });
 ```
 
 ## एपीआई संदर्भ (API Reference)

apps/docs/i18n/hr/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const dohvatnik = vector_index.asRetriever();
 dohvatnik.similarityTopK = 3;
 
 // Dohvati čvorove!
-const čvoroviSaRezultatom = await dohvatnik.retrieve("upitni niz");
+const čvoroviSaRezultatom = await dohvatnik.retrieve({ query: "upitni niz" });
 ```
 
 ## API Referenca

apps/docs/i18n/hu/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Node-ok lekérése!
-const nodesWithScore = await retriever.retrieve("lekérdezési karakterlánc");
+const nodesWithScore = await retriever.retrieve({ query: "lekérdezési karakterlánc" });
 ```
 
 ## API Referencia

apps/docs/i18n/in/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Mengambil node!
-const nodesWithScore = await retriever.retrieve("string query");
+const nodesWithScore = await retriever.retrieve({ query: "string query" });
 ```
 
 ## Referensi API

apps/docs/i18n/it/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Recupera i nodi!
-const nodesWithScore = await retriever.retrieve("stringa di query");
+const nodesWithScore = await retriever.retrieve({ query: "stringa di query" });
 ```
 
 ## Riferimento API

apps/docs/i18n/ja/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // ノードを取得します！
-const nodesWithScore = await retriever.retrieve("クエリ文字列");
+const nodesWithScore = await retriever.retrieve({ query: "クエリ文字列" });
 ```
 
 ## API リファレンス

apps/docs/i18n/ko/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // 노드를 가져옵니다!
-const nodesWithScore = await retriever.retrieve("쿼리 문자열");
+const nodesWithScore = await retriever.retrieve({ query: "쿼리 문자열" });
 ```
 
 ## API 참조

apps/docs/i18n/lt/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const gavėjas = vector_index.asRetriever();
 gavėjas.similarityTopK = 3;
 
 // Išgaunami mazgai!
-const mazgaiSuRezultatu = await gavėjas.retrieve("užklausos eilutė");
+const mazgaiSuRezultatu = await gavėjas.retrieve({ query: "užklausos eilutė" });
 ```
 
 ## API nuorodos (API Reference)

apps/docs/i18n/nl/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Haal knooppunten op!
-const nodesWithScore = await retriever.retrieve("zoekopdracht");
+const nodesWithScore = await retriever.retrieve({ query: "zoekopdracht" });
 ```
 
 ## API Referentie

apps/docs/i18n/no/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Hent noder!
-const nodesWithScore = await retriever.retrieve("spørringsstreng");
+const nodesWithScore = await retriever.retrieve({ query: "spørringsstreng" });
 ```
 
 ## API-referanse

apps/docs/i18n/pl/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Pobierz węzły!
-const nodesWithScore = await retriever.retrieve("ciąg zapytania");
+const nodesWithScore = await retriever.retrieve({ query: "ciąg zapytania" });
 ```
 
 ## Dokumentacja interfejsu API

apps/docs/i18n/pt/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const recuperador = vector_index.asRetriever();
 recuperador.similarityTopK = 3;
 
 // Buscar nós!
-const nósComPontuação = await recuperador.retrieve("string de consulta");
+const nósComPontuação = await recuperador.retrieve({ query: "string de consulta" });
 ```
 
 ## Referência da API

apps/docs/i18n/ro/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const recuperator = vector_index.asRetriever();
 recuperator.similarityTopK = 3;
 
 // Preia nodurile!
-const noduriCuScor = await recuperator.retrieve("șir de interogare");
+const noduriCuScor = await recuperator.retrieve({ query: "șir de interogare" });
 ```
 
 ## Referință API

apps/docs/i18n/ru/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Получение узлов!
-const nodesWithScore = await retriever.retrieve("строка запроса");
+const nodesWithScore = await retriever.retrieve({ query: "строка запроса" });
 ```
 
 ## Справочник по API

apps/docs/i18n/se/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Dohvati čvorove!
-const nodesWithScore = await retriever.retrieve("upitni niz");
+const nodesWithScore = await retriever.retrieve({ query: "upitni niz" });
 ```
 
 ## API Referenca

apps/docs/i18n/sk/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const pridobitelj = vector_index.asRetriever();
 pridobitelj.similarityTopK = 3;
 
 // Pridobivanje vozlišč!
-const vozliščaZRezultatom = await pridobitelj.retrieve("poizvedbeni niz");
+const vozliščaZRezultatom = await pridobitelj.retrieve({ query: "poizvedbeni niz" });
 ```
 
 ## API Sklic

apps/docs/i18n/sl/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Získajte uzly!
-const nodesWithScore = await retriever.retrieve("reťazec dotazu");
+const nodesWithScore = await retriever.retrieve({ query: "reťazec dotazu" });
 ```
 
 ## API Referencia

apps/docs/i18n/sv/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Hämta noder!
-const nodesWithScore = await retriever.retrieve("frågesträng");
+const nodesWithScore = await retriever.retrieve({ query: "frågesträng" });
 ```
 
 ## API-referens

apps/docs/i18n/th/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // เรียกคืนโหนด!
-const nodesWithScore = await retriever.retrieve("query string");
+const nodesWithScore = await retriever.retrieve({ query: "query string" });
 ```
 
 ## API Reference (การอ้างอิง API)

apps/docs/i18n/tr/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Düğümleri getir!
-const nodesWithScore = await retriever.retrieve("sorgu dizesi");
+const nodesWithScore = await retriever.retrieve({ query: "sorgu dizesi" });
 ```
 
 ## API Referansı

apps/docs/i18n/uk/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Отримати вузли!
-const nodesWithScore = await retriever.retrieve("рядок запиту");
+const nodesWithScore = await retriever.retrieve({ query: "рядок запиту" });
 ```
 
 ## Довідник API

apps/docs/i18n/vi/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // Lấy các node!
-const nodesWithScore = await retriever.retrieve("chuỗi truy vấn");
+const nodesWithScore = await retriever.retrieve({ query: "chuỗi truy vấn" });
 ```
 
 ## Tài liệu tham khảo API

apps/docs/i18n/zh-Hans/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -11,7 +11,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // 获取节点！
-const nodesWithScore = await retriever.retrieve("查询字符串");
+const nodesWithScore = await retriever.retrieve({ query: "查询字符串" });
 ```
 
 ## API 参考

apps/docs/i18n/zh_tw/docusaurus-plugin-content-docs/current/modules/low_level/retriever.md

@@ -13,7 +13,7 @@ const retriever = vector_index.asRetriever();
 retriever.similarityTopK = 3;
 
 // 提取節點！
-const nodesWithScore = await retriever.retrieve("查詢字串");
+const nodesWithScore = await retriever.retrieve({ query: "查詢字串" });
 ```
 
 ## API 參考

apps/docs/package.json

@@ -1,6 +1,6 @@
 {
   "name": "docs",
-  "version": "0.0.2",
+  "version": "0.0.4",
   "private": true,
   "scripts": {
     "docusaurus": "docusaurus",
@@ -15,27 +15,28 @@
     "typecheck": "tsc"
   },
   "dependencies": {
-    "@docusaurus/core": "^3.1.1",
-    "@docusaurus/remark-plugin-npm2yarn": "^3.1.1",
-    "@mdx-js/react": "^3.0.0",
+    "@docusaurus/core": "^3.2.1",
+    "@docusaurus/remark-plugin-npm2yarn": "^3.2.1",
+    "@llamaindex/examples": "workspace:*",
+    "@mdx-js/react": "^3.0.1",
     "clsx": "^2.1.0",
-    "postcss": "^8.4.33",
+    "postcss": "^8.4.38",
     "prism-react-renderer": "^2.3.1",
     "raw-loader": "^4.0.2",
     "react": "^18.2.0",
     "react-dom": "^18.2.0"
   },
   "devDependencies": {
-    "@docusaurus/module-type-aliases": "3.1.0",
-    "@docusaurus/preset-classic": "^3.1.1",
-    "@docusaurus/theme-classic": "^3.1.1",
-    "@docusaurus/types": "^3.1.1",
-    "@tsconfig/docusaurus": "^2.0.2",
-    "@types/node": "^18.19.10",
+    "@docusaurus/module-type-aliases": "3.2.0",
+    "@docusaurus/preset-classic": "^3.2.1",
+    "@docusaurus/theme-classic": "^3.2.1",
+    "@docusaurus/types": "^3.2.1",
+    "@tsconfig/docusaurus": "^2.0.3",
+    "@types/node": "^18.19.31",
     "docusaurus-plugin-typedoc": "^0.22.0",
-    "typedoc": "^0.25.7",
+    "typedoc": "^0.25.13",
     "typedoc-plugin-markdown": "^3.17.1",
-    "typescript": "^5.3.3"
+    "typescript": "^5.4.4"
   },
   "browserslist": {
     "production": [

examples/CHANGELOG.md

@@ -0,0 +1,14 @@
+# examples
+
+## 0.0.4
+
+### Patch Changes
+
+- d2e8d0c: add support for Milvus vector store
+- Updated dependencies [d2e8d0c]
+- Updated dependencies [aefc326]
+- Updated dependencies [484a710]
+- Updated dependencies [d766bd0]
+- Updated dependencies [dd95927]
+- Updated dependencies [bf583a7]
+  - llamaindex@0.2.0

examples/Settings.ts

@@ -0,0 +1,29 @@
+import fs from "node:fs/promises";
+
+import { Document, OpenAI, Settings, VectorStoreIndex } from "llamaindex";
+
+Settings.llm = new OpenAI({ model: "gpt-4" });
+
+async function main() {
+  // Load essay from abramov.txt in Node
+  const path = "node_modules/llamaindex/examples/abramov.txt";
+
+  const essay = await fs.readFile(path, "utf-8");
+
+  // Create Document object with essay
+  const document = new Document({ text: essay, id_: path });
+
+  const index = await VectorStoreIndex.fromDocuments([document]);
+
+  // Query the index
+  const queryEngine = index.asQueryEngine();
+
+  const response = await queryEngine.query({
+    query: "What did the author do in college?",
+  });
+
+  // Output response
+  console.log(response.toString());
+}
+
+main().catch(console.error);

examples/agent/multi_document_agent.ts

@@ -6,11 +6,11 @@ import {
   OpenAI,
   OpenAIAgent,
   QueryEngineTool,
+  Settings,
   SimpleNodeParser,
   SimpleToolNodeMapping,
   SummaryIndex,
   VectorStoreIndex,
-  serviceContextFromDefaults,
   storageContextFromDefaults,
 } from "llamaindex";
 
@@ -18,6 +18,8 @@ import { extractWikipedia } from "./helpers/extractWikipedia";
 
 const wikiTitles = ["Brazil", "Canada"];
 
+Settings.llm = new OpenAI({ model: "gpt-4" });
+
 async function main() {
   await extractWikipedia(wikiTitles);
 
@@ -30,11 +32,6 @@ async function main() {
     countryDocs[title] = document;
   }
 
-  const llm = new OpenAI({
-    model: "gpt-4",
-  });
-
-  const serviceContext = serviceContextFromDefaults({ llm });
   const storageContext = await storageContextFromDefaults({
     persistDir: "./storage",
   });
@@ -54,13 +51,11 @@ async function main() {
     console.log(`Creating index for ${title}`);
 
     const vectorIndex = await VectorStoreIndex.init({
-      serviceContext: serviceContext,
       storageContext: storageContext,
       nodes,
     });
 
     const summaryIndex = await SummaryIndex.init({
-      serviceContext: serviceContext,
       nodes,
     });
 
@@ -90,8 +85,7 @@ async function main() {
 
     const agent = new OpenAIAgent({
       tools: queryEngineTools,
-      llm,
-      verbose: true,
+      llm: new OpenAI({ model: "gpt-4" }),
     });
 
     documentAgents[title] = agent;
@@ -126,16 +120,11 @@ async function main() {
     allTools,
     toolMapping,
     VectorStoreIndex,
-    {
-      serviceContext,
-      storageContext,
-    },
   );
 
   const topAgent = new OpenAIAgent({
     toolRetriever: await objectIndex.asRetriever({}),
-    llm,
-    verbose: true,
+    llm: new OpenAI({ model: "gpt-4" }),
     prefixMessages: [
       {
         content:
@@ -154,4 +143,4 @@ async function main() {
   });
 }
 
-main();
+void main();

examples/agent/openai.ts

@@ -1,13 +1,13 @@
 import { FunctionTool, OpenAIAgent } from "llamaindex";
 
 // Define a function to sum two numbers
-function sumNumbers({ a, b }: { a: number; b: number }): number {
-  return a + b;
+function sumNumbers({ a, b }: { a: number; b: number }) {
+  return `${a + b}`;
 }
 
 // Define a function to divide two numbers
-function divideNumbers({ a, b }: { a: number; b: number }): number {
-  return a / b;
+function divideNumbers({ a, b }: { a: number; b: number }) {
+  return `${a / b}`;
 }
 
 // Define the parameters of the sum function as a JSON schema
@@ -24,7 +24,7 @@ const sumJSON = {
     },
   },
   required: ["a", "b"],
-};
+} as const;
 
 const divideJSON = {
   type: "object",
@@ -39,7 +39,7 @@ const divideJSON = {
     },
   },
   required: ["a", "b"],
-};
+} as const;
 
 async function main() {
   // Create a function tool from the sum function
@@ -59,7 +59,6 @@ async function main() {
   // Create an OpenAIAgent with the function tools
   const agent = new OpenAIAgent({
     tools: [functionTool, functionTool2],
-    verbose: true,
   });
 
   // Chat with the agent
@@ -71,6 +70,6 @@ async function main() {
   console.log(String(response));
 }
 
-main().then(() => {
+void main().then(() => {
   console.log("Done");
 });

examples/agent/query_openai_agent.ts

@@ -8,7 +8,7 @@ import {
 async function main() {
   // Load the documents
   const documents = await new SimpleDirectoryReader().loadData({
-    directoryPath: "node_modules/llamaindex/examples/",
+    directoryPath: "node_modules/llamaindex/examples",
   });
 
   // Create a vector index from the documents
@@ -29,7 +29,6 @@ async function main() {
   // Create an OpenAIAgent with the function tools
   const agent = new OpenAIAgent({
     tools: [queryEngineTool],
-    verbose: true,
   });
 
   // Chat with the agent
@@ -41,6 +40,6 @@ async function main() {
   console.log(String(response));
 }
 
-main().then(() => {
+void main().then(() => {
   console.log("Done");
 });

examples/agent/react_agent.ts

@@ -1,13 +1,13 @@
-import { FunctionTool, ReActAgent } from "llamaindex";
+import { Anthropic, FunctionTool, ReActAgent } from "llamaindex";
 
 // Define a function to sum two numbers
-function sumNumbers({ a, b }: { a: number; b: number }): number {
-  return a + b;
+function sumNumbers({ a, b }: { a: number; b: number }) {
+  return `${a + b}`;
 }
 
 // Define a function to divide two numbers
-function divideNumbers({ a, b }: { a: number; b: number }): number {
-  return a / b;
+function divideNumbers({ a, b }: { a: number; b: number }) {
+  return `${a / b}`;
 }
 
 // Define the parameters of the sum function as a JSON schema
@@ -24,7 +24,7 @@ const sumJSON = {
     },
   },
   required: ["a", "b"],
-};
+} as const;
 
 const divideJSON = {
   type: "object",
@@ -39,7 +39,7 @@ const divideJSON = {
     },
   },
   required: ["a", "b"],
-};
+} as const;
 
 async function main() {
   // Create a function tool from the sum function
@@ -56,10 +56,15 @@ async function main() {
     parameters: divideJSON,
   });
 
-  // Create an OpenAIAgent with the function tools
+  const anthropic = new Anthropic({
+    apiKey: process.env.ANTHROPIC_API_KEY,
+    model: "claude-3-opus",
+  });
+
+  // Create an ReActAgent with the function tools
   const agent = new ReActAgent({
+    llm: anthropic,
     tools: [functionTool, functionTool2],
-    verbose: true,
   });
 
   // Chat with the agent
@@ -71,6 +76,6 @@ async function main() {
   console.log(String(response));
 }
 
-main().then(() => {
+void main().then(() => {
   console.log("Done");
 });

examples/agent/step_wise_openai.ts

@@ -0,0 +1,94 @@
+import { FunctionTool, OpenAIAgent } from "llamaindex";
+
+// Define a function to sum two numbers
+function sumNumbers({ a, b }: { a: number; b: number }) {
+  return `${a + b}`;
+}
+
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }) {
+  return `${a / b}`;
+}
+
+// Define the parameters of the sum function as a JSON schema
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
+    },
+  },
+  required: ["a", "b"],
+} as const;
+
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The dividend",
+    },
+    b: {
+      type: "number",
+      description: "The divisor",
+    },
+  },
+  required: ["a", "b"],
+} as const;
+
+async function main() {
+  // Create a function tool from the sum function
+  const functionTool = new FunctionTool(sumNumbers, {
+    name: "sumNumbers",
+    description: "Use this function to sum two numbers",
+    parameters: sumJSON,
+  });
+
+  // Create a function tool from the divide function
+  const functionTool2 = new FunctionTool(divideNumbers, {
+    name: "divideNumbers",
+    description: "Use this function to divide two numbers",
+    parameters: divideJSON,
+  });
+
+  // Create an OpenAIAgent with the function tools
+  const agent = new OpenAIAgent({
+    tools: [functionTool, functionTool2],
+  });
+
+  // Create a task to sum and divide numbers
+  const task = agent.createTask("How much is 5 + 5? then divide by 2");
+
+  let count = 0;
+
+  while (true) {
+    const stepOutput = await agent.runStep(task.taskId);
+
+    console.log(`Runnning step ${count++}`);
+    console.log(`======== OUTPUT ==========`);
+    if (stepOutput.output.response) {
+      console.log(stepOutput.output.response);
+    } else {
+      console.log(stepOutput.output.sources);
+    }
+    console.log(`==========================`);
+
+    if (stepOutput.isLast) {
+      const finalResponse = await agent.finalizeResponse(
+        task.taskId,
+        stepOutput,
+      );
+      console.log({ finalResponse });
+      break;
+    }
+  }
+}
+
+void main().then(() => {
+  console.log("Done");
+});

examples/agent/step_wise_query_tool.ts

@@ -8,7 +8,7 @@ import {
 async function main() {
   // Load the documents
   const documents = await new SimpleDirectoryReader().loadData({
-    directoryPath: "node_modules/llamaindex/examples/",
+    directoryPath: "node_modules/llamaindex/examples",
   });
 
   // Create a vector index from the documents
@@ -29,18 +29,35 @@ async function main() {
   // Create an OpenAIAgent with the function tools
   const agent = new OpenAIAgent({
     tools: [queryEngineTool],
-    verbose: true,
   });
 
-  // Chat with the agent
-  const response = await agent.chat({
-    message: "What was his salary?",
-  });
+  const task = agent.createTask("What was his salary?");
 
-  // Print the response
-  console.log(String(response));
+  let count = 0;
+
+  while (true) {
+    const stepOutput = await agent.runStep(task.taskId);
+
+    console.log(`Runnning step ${count++}`);
+    console.log(`======== OUTPUT ==========`);
+    if (stepOutput.output.response) {
+      console.log(stepOutput.output.response);
+    } else {
+      console.log(stepOutput.output.sources);
+    }
+    console.log(`==========================`);
+
+    if (stepOutput.isLast) {
+      const finalResponse = await agent.finalizeResponse(
+        task.taskId,
+        stepOutput,
+      );
+      console.log({ finalResponse });
+      break;
+    }
+  }
 }
 
-main().then(() => {
+void main().then(() => {
   console.log("Done");
 });

examples/agent/step_wise_react.ts

@@ -0,0 +1,89 @@
+import { FunctionTool, ReActAgent } from "llamaindex";
+
+// Define a function to sum two numbers
+function sumNumbers({ a, b }: { a: number; b: number }) {
+  return `${a + b}`;
+}
+
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }) {
+  return `${a / b}`;
+}
+
+// Define the parameters of the sum function as a JSON schema
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
+    },
+  },
+  required: ["a", "b"],
+} as const;
+
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The dividend",
+    },
+    b: {
+      type: "number",
+      description: "The divisor",
+    },
+  },
+  required: ["a", "b"],
+} as const;
+
+async function main() {
+  // Create a function tool from the sum function
+  const functionTool = new FunctionTool(sumNumbers, {
+    name: "sumNumbers",
+    description: "Use this function to sum two numbers",
+    parameters: sumJSON,
+  });
+
+  // Create a function tool from the divide function
+  const functionTool2 = new FunctionTool(divideNumbers, {
+    name: "divideNumbers",
+    description: "Use this function to divide two numbers",
+    parameters: divideJSON,
+  });
+
+  // Create an OpenAIAgent with the function tools
+  const agent = new ReActAgent({
+    tools: [functionTool, functionTool2],
+  });
+
+  const task = agent.createTask("Divide 16 by 2 then add 20");
+
+  let count = 0;
+
+  while (true) {
+    const stepOutput = await agent.runStep(task.taskId);
+
+    console.log(`Runnning step ${count++}`);
+    console.log(`======== OUTPUT ==========`);
+    console.log(stepOutput.output);
+    console.log(`==========================`);
+
+    if (stepOutput.isLast) {
+      const finalResponse = await agent.finalizeResponse(
+        task.taskId,
+        stepOutput,
+      );
+      console.log({ finalResponse });
+      break;
+    }
+  }
+}
+
+void main().then(() => {
+  console.log("Done");
+});

examples/agent/stream_openai_agent.ts

@@ -0,0 +1,76 @@
+import { FunctionTool, OpenAIAgent } from "llamaindex";
+
+// Define a function to sum two numbers
+function sumNumbers({ a, b }: { a: number; b: number }) {
+  return `${a + b}`;
+}
+
+// Define a function to divide two numbers
+function divideNumbers({ a, b }: { a: number; b: number }) {
+  return `${a / b}`;
+}
+
+// Define the parameters of the sum function as a JSON schema
+const sumJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The first number",
+    },
+    b: {
+      type: "number",
+      description: "The second number",
+    },
+  },
+  required: ["a", "b"],
+} as const;
+
+const divideJSON = {
+  type: "object",
+  properties: {
+    a: {
+      type: "number",
+      description: "The dividend",
+    },
+    b: {
+      type: "number",
+      description: "The divisor",
+    },
+  },
+  required: ["a", "b"],
+} as const;
+
+async function main() {
+  // Create a function tool from the sum function
+  const functionTool = FunctionTool.from(sumNumbers, {
+    name: "sumNumbers",
+    description: "Use this function to sum two numbers",
+    parameters: sumJSON,
+  });
+
+  // Create a function tool from the divide function
+  const functionTool2 = FunctionTool.from(divideNumbers, {
+    name: "divideNumbers",
+    description: "Use this function to divide two numbers",
+    parameters: divideJSON,
+  });
+
+  // Create an OpenAIAgent with the function tools
+  const agent = new OpenAIAgent({
+    tools: [functionTool, functionTool2],
+  });
+
+  const stream = await agent.chat({
+    message: "Divide 16 by 2 then add 20",
+    stream: true,
+  });
+
+  for await (const chunk of stream.response) {
+    process.stdout.write(chunk.response);
+  }
+}
+
+void main().then(() => {
+  console.log("\nDone");
+});

examples/agent/wiki.ts

@@ -0,0 +1,27 @@
+import { OpenAI, OpenAIAgent, WikipediaTool } from "llamaindex";
+
+async function main() {
+  const llm = new OpenAI({ model: "gpt-4-turbo" });
+  const wikiTool = new WikipediaTool();
+
+  // Create an OpenAIAgent with the Wikipedia tool
+  const agent = new OpenAIAgent({
+    llm,
+    tools: [wikiTool],
+  });
+
+  // Chat with the agent
+  const response = await agent.chat({
+    message: "Who was Goethe?",
+    stream: true,
+  });
+
+  for await (const chunk of response.response) {
+    process.stdout.write(chunk.response);
+  }
+}
+
+(async function () {
+  await main();
+  console.log("\nDone");
+})();

examples/anthropic/chat.ts

@@ -3,6 +3,7 @@ import { Anthropic } from "llamaindex";
 (async () => {
   const anthropic = new Anthropic({
     apiKey: process.env.ANTHROPIC_API_KEY,
+    model: "claude-3-opus",
   });
   const result = await anthropic.chat({
     messages: [

examples/anthropic/chat_interactive.ts

@@ -0,0 +1,34 @@
+import { Anthropic, SimpleChatEngine, SimpleChatHistory } from "llamaindex";
+import { stdin as input, stdout as output } from "node:process";
+import readline from "node:readline/promises";
+
+(async () => {
+  const llm = new Anthropic({
+    apiKey: process.env.ANTHROPIC_API_KEY,
+    model: "claude-3-opus",
+  });
+  // chatHistory will store all the messages in the conversation
+  const chatHistory = new SimpleChatHistory({
+    messages: [
+      {
+        content: "You want to talk in rhymes.",
+        role: "system",
+      },
+    ],
+  });
+  const chatEngine = new SimpleChatEngine({
+    llm,
+    chatHistory,
+  });
+  const rl = readline.createInterface({ input, output });
+
+  while (true) {
+    const query = await rl.question("User: ");
+    process.stdout.write("Assistant: ");
+    const stream = await chatEngine.chat({ message: query, stream: true });
+    for await (const chunk of stream) {
+      process.stdout.write(chunk.response);
+    }
+    process.stdout.write("\n");
+  }
+})();