Release 0.11.13 (#2060 )

chore: bump sdk version (#2062 )
chore: re-add lib folder from docs and rename it to libs (so pnpm clean doesn't delete it)
2026-07-02 20:13:52 -04:00 · 2025-07-07 11:24:21 +07:00 · 2025-07-03 12:05:16 -07:00 · 2025-07-03 11:03:05 +07:00 · 2025-07-01 22:53:47 -07:00 · 2025-07-02 11:41:55 +07:00
228 changed files with 17607 additions and 1059 deletions
@@ -25,7 +25,7 @@ Make sure you have Node.js LTS (Long-term Support) installed. You can check your

 ```shell
 node -v
-# v20.x.x
+# v22.x.x
 ```

 ### Use pnpm
@@ -1,5 +1,28 @@
 # @llamaindex/doc

+## 0.2.33
+
+### Patch Changes
+
+- Updated dependencies [47a7555]
+  - @llamaindex/cloud@4.0.18
+  - llamaindex@0.11.13
+
+## 0.2.32
+
+### Patch Changes
+
+- Updated dependencies [d578889]
+- Updated dependencies [0fcc92f]
+- Updated dependencies [515a8b9]
+  - @llamaindex/core@0.6.13
+  - llamaindex@0.11.12
+  - @llamaindex/cloud@4.0.17
+  - @llamaindex/node-parser@2.0.13
+  - @llamaindex/openai@0.4.7
+  - @llamaindex/readers@3.1.12
+  - @llamaindex/workflow@1.1.13
+
 ## 0.2.31

 ### Patch Changes
@@ -3,6 +3,8 @@
 This is a Next.js application generated with
 [Create Fumadocs](https://github.com/fuma-nama/fumadocs).

+> Note: Before running the development server, make sure to build the whole project first, see [CONTRIBUTING.md](../../CONTRIBUTING.md) for more details.
+
 Run development server:

 ```bash
@@ -12,9 +12,9 @@
  },
  "aliases": {
    "components": "@/components",
-    "utils": "@/lib/utils",
+    "utils": "@/libs/utils",
    "ui": "@/components/ui",
-    "lib": "@/lib",
+    "lib": "@/libs",
    "hooks": "@/hooks"
  }
 }
@@ -1,6 +1,6 @@
 {
  "name": "@llamaindex/doc",
-  "version": "0.2.31",
+  "version": "0.2.33",
  "private": true,
  "scripts": {
    "postinstall": "fumadocs-mdx",
@@ -15,7 +15,6 @@
  "dependencies": {
    "@huggingface/transformers": "^3.5.0",
    "@icons-pack/react-simple-icons": "^10.1.0",
-    "@llamaindex/workflow-docs": "0.1.1",
    "@llamaindex/chat-ui-docs": "^0.0.5",
    "@llamaindex/cloud": "workspace:*",
    "@llamaindex/core": "workspace:*",
@@ -23,8 +22,10 @@
    "@llamaindex/openai": "workspace:*",
    "@llamaindex/readers": "workspace:*",
    "@llamaindex/workflow": "workspace:*",
+    "@llamaindex/workflow-docs": "0.1.1",
    "@mdx-js/mdx": "^3.1.0",
    "@monaco-editor/react": "^4.7.0",
+    "@next/third-parties": "^15.3.4",
    "@number-flow/react": "^0.3.4",
    "@radix-ui/react-dialog": "^1.1.2",
    "@radix-ui/react-icons": "^1.3.2",
@@ -10,7 +10,7 @@ import { MagicMove } from "@/components/magic-move";
 import { NpmInstall } from "@/components/npm-install";
 import { Supports } from "@/components/supports";
 import { Button } from "@/components/ui/button";
-import { DOCUMENT_URL } from "@/lib/const";
+import { DOCUMENT_URL } from "@/libs/const";
 import { SiStackblitz } from "@icons-pack/react-simple-icons";
 import { Blocks, Bot, Footprints, Terminal } from "lucide-react";
 import Link from "next/link";
@@ -1,4 +1,4 @@
-import { source } from "@/lib/source";
+import { source } from "@/libs/source";
 import { structure } from "fumadocs-core/mdx-plugins";
 import { createFromSource } from "fumadocs-core/search/server";

@@ -1,6 +1,6 @@
 import * as demos from "@/components/demo/lazy";
-import { createMetadata, metadataImage } from "@/lib/metadata";
-import { openapi, source } from "@/lib/source";
+import { createMetadata, metadataImage } from "@/libs/metadata";
+import { openapi, source } from "@/libs/source";
 import * as Icons from "@icons-pack/react-simple-icons";
 import { APIPage } from "fumadocs-openapi/ui";
 import { Popup, PopupContent, PopupTrigger } from "fumadocs-twoslash/ui";
@@ -1,5 +1,5 @@
 import { baseOptions } from "@/app/layout.config";
-import { source } from "@/lib/source";
+import { source } from "@/libs/source";
 import "fumadocs-twoslash/twoslash.css";
 import { DocsLayout } from "fumadocs-ui/layouts/docs";
 import type { ReactNode } from "react";
@@ -1,4 +1,4 @@
-import { DOCUMENT_URL } from "@/lib/const";
+import { DOCUMENT_URL } from "@/libs/const";
 import type { BaseLayoutProps } from "fumadocs-ui/layouts/shared";
 import Image from "next/image";

@@ -1,5 +1,6 @@
 import { AIProvider } from "@/actions";
 import { TooltipProvider } from "@/components/ui/tooltip";
+import { GoogleAnalytics } from "@next/third-parties/google";
 import { RootProvider } from "fumadocs-ui/provider";
 import { Inter } from "next/font/google";
 import type { ReactNode } from "react";
@@ -39,6 +40,7 @@ export default function Layout({ children }: { children: ReactNode }) {
          </AIProvider>
        </TooltipProvider>
      </body>
+      <GoogleAnalytics gaId="G-NB9B8LW9W5" />
    </html>
  );
 }
@@ -1,5 +1,5 @@
 import { generateOGImage } from "@/app/og/[...slug]/og";
-import { metadataImage } from "@/lib/metadata";
+import { metadataImage } from "@/libs/metadata";
 import { type ImageResponse } from "next/og";
 import { readFileSync } from "node:fs";

@@ -1,6 +1,6 @@
 import ContributorCounter from "@/components/contributor-count";
 import { buttonVariants } from "@/components/ui/button";
-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";
 import { Heart } from "lucide-react";
 import { ReactElement } from "react";

@@ -1,5 +1,5 @@
-import { fetchContributors } from "@/lib/get-contributors";
-import { cn } from "@/lib/utils";
+import { fetchContributors } from "@/libs/get-contributors";
+import { cn } from "@/libs/utils";
 import Image from "next/image";
 import type { HTMLAttributes, ReactElement } from "react";

@@ -1,5 +1,5 @@
 "use client";
-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";
 import { TerminalIcon } from "lucide-react";
 import {
  Fragment,
@@ -1,4 +1,4 @@
-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";
 import { LucideIcon } from "lucide-react";
 import { HTMLAttributes, ReactElement, ReactNode } from "react";

@@ -1,6 +1,6 @@
 "use client";
 import { Button } from "@/components/ui/button";
-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";
 import { CodeBlock } from "fumadocs-ui/components/codeblock";
 import { RotateCcw } from "lucide-react";
 import { useTheme } from "next-themes";
@@ -1,6 +1,6 @@
 "use client";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";
 import Image from "next/image";
 import { ReactNode } from "react";
 import { IconAI, IconUser } from "./ui/icons";
@@ -1,4 +1,4 @@
-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";
 import {
  AnimatePresence,
  motion,
@@ -1,7 +1,7 @@
 import { cva, type VariantProps } from "class-variance-authority";
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 const alertVariants = cva(
  "relative w-full rounded-lg border px-4 py-3 text-sm [&>svg+div]:translate-y-[-3px] [&>svg]:absolute [&>svg]:left-4 [&>svg]:top-4 [&>svg]:text-foreground [&>svg~*]:pl-7",
@@ -1,7 +1,7 @@
 import { cva, type VariantProps } from "class-variance-authority";
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 const badgeVariants = cva(
  "inline-flex items-center rounded-md border px-2.5 py-0.5 text-xs font-semibold transition-colors focus:outline-none focus:ring-2 focus:ring-ring focus:ring-offset-2",
@@ -2,7 +2,7 @@ import { Slot } from "@radix-ui/react-slot";
 import { cva, type VariantProps } from "class-variance-authority";
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 const buttonVariants = cva(
  "inline-flex items-center justify-center gap-2 whitespace-nowrap rounded-md text-sm font-medium transition-colors focus-visible:outline-none focus-visible:ring-1 focus-visible:ring-ring disabled:pointer-events-none disabled:opacity-50 [&_svg]:pointer-events-none [&_svg]:size-4 [&_svg]:shrink-0",
@@ -4,7 +4,7 @@ import * as DialogPrimitive from "@radix-ui/react-dialog";
 import { Cross2Icon } from "@radix-ui/react-icons";
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 const Dialog = DialogPrimitive.Root;

@@ -1,4 +1,4 @@
-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 export function IconAI({ className, ...props }: React.ComponentProps<"svg">) {
  return (
@@ -1,5 +1,5 @@
 "use client";
-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";
 import { animate, motion, useMotionValue } from "framer-motion";
 import { useEffect, useState } from "react";
 import useMeasure from "react-use-measure";
@@ -1,6 +1,6 @@
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 export type InputProps = React.InputHTMLAttributes<HTMLInputElement>;

@@ -4,7 +4,7 @@ import * as LabelPrimitive from "@radix-ui/react-label";
 import { cva, type VariantProps } from "class-variance-authority";
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 const labelVariants = cva(
  "text-sm font-medium leading-none peer-disabled:cursor-not-allowed peer-disabled:opacity-70",
@@ -1,4 +1,4 @@
-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 function Skeleton({
  className,
@@ -3,7 +3,7 @@
 import * as SliderPrimitive from "@radix-ui/react-slider";
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 const Slider = React.forwardRef<
  React.ElementRef<typeof SliderPrimitive.Root>,
@@ -1,6 +1,6 @@
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 export type TextareaProps = React.TextareaHTMLAttributes<HTMLTextAreaElement>;

@@ -3,7 +3,7 @@
 import * as TooltipPrimitive from "@radix-ui/react-tooltip";
 import * as React from "react";

-import { cn } from "@/lib/utils";
+import { cn } from "@/libs/utils";

 const TooltipProvider = TooltipPrimitive.Provider;

@@ -0,0 +1,182 @@
+---
+title: Memory
+description: Manage conversation history and context with agents
+---
+
+## Concept
+
+Memory is a core component of agentic systems. It allows you to store and retrieve information from the past.
+
+In LlamaIndexTS, you can create memory by using the `createMemory` function. This function will return a `Memory` object, which you can then use to store and retrieve information.
+
+As the agent runs, it will make calls to `add()` to store information, and `get()` to retrieve information. 
+
+## Usage
+
+A `Memory` object has both short-term memory (i.e. a FIFO queue of messages) and optionally long-term memory (i.e. extracting information over time).
+
+`get()` always returns all messages stored in the memory. The longer the agent runs, this will exceed the context window of the agent. To avoid this, the agent is using the `getLLM` method to get the last X messages that fit into the context window.
+
+### Configuring Memory for an Agent
+
+Here we're creating a memory with a static block (read more about [memory blocks](#long-term-memory)) that contains some information about the user.
+
+```ts twoslash
+import { openai } from "@llamaindex/openai";
+import { agent } from "@llamaindex/workflow";
+import { createMemory, staticBlock } from "llamaindex";
+
+const llm = openai({ model: "gpt-4.1-mini" });
+
+// Create memory with predefined context
+const memory = createMemory({
+  memoryBlocks: [
+    staticBlock({
+      content:
+        "The user is a software engineer who loves TypeScript and LlamaIndex.",
+    }),
+  ],
+});
+
+// Create an agent with the memory
+const workflow = agent({
+  name: "assistant",
+  llm,
+  memory,
+});
+
+const result = await workflow.run("What is my name?");
+console.log("Response:", result.data.result);
+```
+
+### Using Vercel format
+
+You can also put messages in Vercel format directly to the memory:
+
+```ts
+await memory.add({
+  id: "1",
+  createdAt: new Date(),
+  role: "user",
+  content: "Hello!",
+  options: {
+    parts: [
+      {
+        type: "file",
+        data: "base64...",
+        mimeType: "image/png",
+      },
+    ],
+  },
+});
+```
+
+If you call `get`, messages are usually retrieved in the LlamaIndexTS format (type `ChatMessage`). If you specify the `type` parameter using `get`, you can return the messages in different formats. E.g.: using `type: "vercel"`, you can return the messages in Vercel format:
+
+```ts
+const messages = await memory.get({ type: "vercel" });
+console.log(messages);
+```
+
+## Customizing Memory
+
+### Short-Term Memory
+
+The `Memory` object will store all the messages that are added to the `Memory` object. Unless you call `clear()`, no messages are removed from the memory. This is the short-term memory (usually you will store the memory of one user session there) which is augmented by the long-term memory.
+
+Calling `getLLM` will retrieve messages from long-term memory and ensure that the given `tokenLimit` is not reached. These are the messages that you will sent to the LLM.
+
+For initialization, you call `createMemory` with the following options:
+
+- `tokenLimit`: Maximum tokens for memory retrieval using `getLLM` (default: 30000).
+- `shortTermTokenLimitRatio`: Ratio of tokens for short-term vs long-term memory (default: 0.7)
+- `customAdapters`: Custom message adapters for different message formats. LlamaIndex (`ChatMessageAdapter`) and Vercel (`VercelMessageAdapter`) are built-in adapters.
+- `memoryBlocks`: Memory blocks for long-term storage, see [Long-Term Memory](#long-term-memory)
+
+Example:
+
+```ts
+const memory = createMemory({
+    tokenLimit=40000,
+    shortTermTokenLimitRatio=0.5,
+});
+```
+
+### Long-Term Memory
+
+Long-term memory is represented as `Memory Block` objects. These objects contain information that are from previous user sessions or from the beginning of the current conversation. When memory is retrieved (by calling `getLLM`), the short-term and long-term memories are merged together within the given `tokenLimit`. 
+
+Currently, there are two predefined memory blocks:
+
+- `staticBlock`: A memory block that stores a static piece of information.
+- `factExtractionBlock`: A memory block that extracts facts from the chat history.
+
+This sounds a bit complicated, but it's actually quite simple. Let's look at an example:
+
+```ts
+import { createMemory, factExtractionBlock, staticBlock } from "llamaindex";
+
+const memoryBlocks= [
+  staticBlock({
+    id: "core_info",
+    content: "My name is Logan, and I live in Saskatoon. I work at LlamaIndex.",
+  }),
+  factExtractionBlock({
+    id: "user-extracted_info",
+    priority: 1,
+    llm: llm,
+    maxFacts: 50,
+  }),
+];
+```
+
+Here, we've setup two memory blocks:
+
+- `core_info`: A static memory block that stores some core information about the user. This information will always be inserted into the memory. The type used is `MessageContent` to support multi-modal content.
+- `extracted_info`: An extracted memory block that will extract information from the chat history. Here we've passed in the `llm` to use to extract facts from the chat history, and set the `maxFacts` to 50. If the number of extracted facts exceeds this limit, the `maxFacts` will be automatically summarized and reduced to leave room for new information.
+
+You'll also notice that we've set the `priority` for the `factExtractionBlock` block. This is used to determine the handling when the memory blocks content (i.e. long-term memory) + short-term memory exceeds the token limit on the `Memory` object.
+
+- `priority=0`: This block will always be kept in memory (`staticBlocks` always have priority 0.)
+- `priority=1, 2, 3, etc`: This determines the order in which memory blocks are truncated when the memory exceeds the token limit, to help the overall short-term memory + long-term memory content be less than or equal to the `tokenLimit`.
+
+Now, let's pass these blocks into the `createMemory` function:
+
+```ts
+const memory = createMemory({
+  tokenLimit: 40000,
+  memoryBlocks: memoryBlocks,
+)
+```
+
+When memory is retrieved (using `getLLM`), the short-term and long-term memories are merged together. The `Memory` object will ensure that the short-term memory + long-term memory content is less than or equal to the `tokenLimit`. If it is longer, messages are retrieved in the following order:
+
+1. StaticMemoryBlock (information always included)
+2. LongTermMemoryBlock (depending on priority)
+3. ShortTermMemoryBlock 
+4. Transient messages
+
+The amount of short-term memory included is specified by the `shortTermTokenLimitRatio`. If it's set to `0.7`, 70% of the `tokenLimit` is used for short-term memory (not including the static memory block).
+
+## Persistence with Snapshots
+
+Save and restore memory state:
+
+```ts twoslash
+import { createMemory, loadMemory } from "llamaindex";
+
+const memory = createMemory();
+
+// Add some messages
+await memory.add({ role: "user", content: "Hello!" });
+
+// Create snapshot
+const snapshot = memory.snapshot();
+
+// Later, restore from the snapshot
+const restoredMemory = loadMemory(snapshot);
+```
+
+## Examples
+
+Want to learn more about the Memory class? Check out our example codes in [Github](https://github.com/run-llama/LlamaIndexTS/tree/main/examples/agents/memory).
@@ -1,4 +1,11 @@
 {
  "title": "Data",
-  "pages": ["index", "readers", "data_index", "ingestion_pipeline", "stores"]
+  "pages": [
+    "index",
+    "memory",
+    "readers",
+    "data_index",
+    "ingestion_pipeline",
+    "stores"
+  ]
 }
@@ -11,6 +11,7 @@ A retriever in LlamaIndex is what is used to fetch `Node`s from an index using a
 - [KeywordTableLLMRetriever](/docs/api/classes/KeywordTableLLMRetriever) uses an LLM to extract keywords from the query and retrieve relevant nodes based on keyword matches.
 - [KeywordTableSimpleRetriever](/docs/api/classes/KeywordTableSimpleRetriever) uses a basic frequency-based approach to extract keywords and retrieve nodes.
 - [KeywordTableRAKERetriever](/docs/api/classes/KeywordTableRAKERetriever) uses the RAKE (Rapid Automatic Keyword Extraction) algorithm to extract keywords from the query, focusing on co-occurrence and context for keyword-based retrieval.
+- [Bm25Retriever](/docs/api/classes/Bm25Retriever) uses the BM25 algorithm to extract keywords from the query and retrieve relevant nodes based on keyword matches.

 ```typescript
 const retriever = vectorIndex.asRetriever({
@@ -1,30 +0,0 @@
-import { createMetadataImage } from 'fumadocs-core/server';
-import { source } from '@/lib/source';
-import { Metadata } from 'next';
-
-export const metadataImage = createMetadataImage({
-	source,
-	imageRoute: 'og',
-});
-
-export function createMetadata(override: Metadata): Metadata {
-	return {
-		...override,
-		openGraph: {
-			title: override.title ?? undefined,
-			description: override.description ?? undefined,
-			url: 'https://ts.llamaindex.ai/',
-			images: '/og.png',
-			siteName: 'LlamaIndex.TS',
-			...override.openGraph,
-		},
-		twitter: {
-			card: 'summary_large_image',
-			creator: '@llama_index',
-			title: override.title ?? undefined,
-			description: override.description ?? undefined,
-			images: '/og.png',
-			...override.twitter,
-		},
-	};
-}
@@ -1,6 +0,0 @@
-import { clsx, type ClassValue } from "clsx"
-import { twMerge } from "tailwind-merge"
-
-export function cn(...inputs: ClassValue[]) {
-  return twMerge(clsx(inputs))
-}
@@ -1,2 +1,2 @@
 // when we are ready, change to /docs/llamaindex
-export const DOCUMENT_URL = '/docs/llamaindex'
+export const DOCUMENT_URL = "/docs/llamaindex";
@@ -10,7 +10,7 @@ export async function fetchContributors(
 ): Promise<Contributor[]> {
  const headers = new Headers();
  if (process.env.GITHUB_TOKEN)
-    headers.set('Authorization', `Bearer ${process.env.GITHUB_TOKEN}`);
+    headers.set("Authorization", `Bearer ${process.env.GITHUB_TOKEN}`);

  const response = await fetch(
    `https://api.github.com/repos/${repoOwner}/${repoName}/contributors?per_page=50`,
@@ -26,6 +26,6 @@ export async function fetchContributors(

  const contributors = (await response.json()) as Contributor[];
  return contributors
-    .filter((contributor) => !contributor.login.endsWith('[bot]'))
+    .filter((contributor) => !contributor.login.endsWith("[bot]"))
    .sort((a, b) => b.contributions - a.contributions);
 }
@@ -0,0 +1,30 @@
+import { source } from "@/libs/source";
+import { createMetadataImage } from "fumadocs-core/server";
+import { Metadata } from "next";
+
+export const metadataImage = createMetadataImage({
+  source,
+  imageRoute: "og",
+});
+
+export function createMetadata(override: Metadata): Metadata {
+  return {
+    ...override,
+    openGraph: {
+      title: override.title ?? undefined,
+      description: override.description ?? undefined,
+      url: "https://ts.llamaindex.ai/",
+      images: "/og.png",
+      siteName: "LlamaIndex.TS",
+      ...override.openGraph,
+    },
+    twitter: {
+      card: "summary_large_image",
+      creator: "@llama_index",
+      title: override.title ?? undefined,
+      description: override.description ?? undefined,
+      images: "/og.png",
+      ...override.twitter,
+    },
+  };
+}
@@ -1,9 +1,9 @@
-import { docs } from '@/.source';
-import { loader } from 'fumadocs-core/source';
+import { docs } from "@/.source";
+import { loader } from "fumadocs-core/source";
 import { createOpenAPI } from "fumadocs-openapi/server";

 export const source = loader({
-  baseUrl: '/docs',
+  baseUrl: "/docs",
  source: docs.toFumadocsSource(),
 });

@@ -0,0 +1,6 @@
+import { clsx, type ClassValue } from "clsx";
+import { twMerge } from "tailwind-merge";
+
+export function cn(...inputs: ClassValue[]) {
+  return twMerge(clsx(inputs));
+}
@@ -1,5 +1,18 @@
 # @llamaindex/cloudflare-worker-agent-test

+## 0.0.174
+
+### Patch Changes
+
+- llamaindex@0.11.13
+
+## 0.0.173
+
+### Patch Changes
+
+- Updated dependencies [515a8b9]
+  - llamaindex@0.11.12
+
 ## 0.0.172

 ### Patch Changes
@@ -1,6 +1,6 @@
 {
  "name": "@llamaindex/cloudflare-worker-agent-test",
-  "version": "0.0.172",
+  "version": "0.0.174",
  "type": "module",
  "private": true,
  "scripts": {
@@ -1,5 +1,18 @@
 # @llamaindex/llama-parse-browser-test

+## 0.0.73
+
+### Patch Changes
+
+- Updated dependencies [47a7555]
+  - @llamaindex/cloud@4.0.18
+
+## 0.0.72
+
+### Patch Changes
+
+- @llamaindex/cloud@4.0.17
+
 ## 0.0.71

 ### Patch Changes
@@ -1,7 +1,7 @@
 {
  "name": "@llamaindex/llama-parse-browser-test",
  "private": true,
-  "version": "0.0.71",
+  "version": "0.0.73",
  "type": "module",
  "scripts": {
    "dev": "vite",
@@ -1,5 +1,18 @@
 # @llamaindex/next-agent-test

+## 0.1.174
+
+### Patch Changes
+
+- llamaindex@0.11.13
+
+## 0.1.173
+
+### Patch Changes
+
+- Updated dependencies [515a8b9]
+  - llamaindex@0.11.12
+
 ## 0.1.172

 ### Patch Changes
@@ -1,6 +1,6 @@
 {
  "name": "@llamaindex/next-agent-test",
-  "version": "0.1.172",
+  "version": "0.1.174",
  "private": true,
  "scripts": {
    "dev": "next dev",
@@ -1,5 +1,18 @@
 # test-edge-runtime

+## 0.1.173
+
+### Patch Changes
+
+- llamaindex@0.11.13
+
+## 0.1.172
+
+### Patch Changes
+
+- Updated dependencies [515a8b9]
+  - llamaindex@0.11.12
+
 ## 0.1.171

 ### Patch Changes
@@ -1,6 +1,6 @@
 {
  "name": "@llamaindex/nextjs-edge-runtime-test",
-  "version": "0.1.171",
+  "version": "0.1.173",
  "private": true,
  "scripts": {
    "dev": "next dev",
@@ -1,5 +1,20 @@
 # @llamaindex/next-node-runtime

+## 0.1.42
+
+### Patch Changes
+
+- llamaindex@0.11.13
+
+## 0.1.41
+
+### Patch Changes
+
+- Updated dependencies [515a8b9]
+  - llamaindex@0.11.12
+  - @llamaindex/huggingface@0.1.17
+  - @llamaindex/readers@3.1.12
+
 ## 0.1.40

 ### Patch Changes
@@ -1,6 +1,6 @@
 {
  "name": "@llamaindex/next-node-runtime-test",
-  "version": "0.1.40",
+  "version": "0.1.42",
  "private": true,
  "scripts": {
    "dev": "next dev",
@@ -1,5 +1,18 @@
 # vite-import-llamaindex

+## 0.0.40
+
+### Patch Changes
+
+- llamaindex@0.11.13
+
+## 0.0.39
+
+### Patch Changes
+
+- Updated dependencies [515a8b9]
+  - llamaindex@0.11.12
+
 ## 0.0.38

 ### Patch Changes
@@ -1,7 +1,7 @@
 {
  "name": "vite-import-llamaindex",
  "private": true,
-  "version": "0.0.38",
+  "version": "0.0.40",
  "type": "module",
  "scripts": {
    "build": "vite build",
@@ -1,9 +1 @@
-{
-  "root": [
-    "./src/main.ts",
-    "./vite.config.ts",
-    "./tsconfig.json"
-  ],
-  "errors": true,
-  "version": "5.7.3"
-}
+{"root":["./src/main.ts","./vite.config.ts"],"version":"5.7.3"}
@@ -1,5 +1,18 @@
 # @llamaindex/waku-query-engine-test

+## 0.0.174
+
+### Patch Changes
+
+- llamaindex@0.11.13
+
+## 0.0.173
+
+### Patch Changes
+
+- Updated dependencies [515a8b9]
+  - llamaindex@0.11.12
+
 ## 0.0.172

 ### Patch Changes
@@ -1,6 +1,6 @@
 {
  "name": "@llamaindex/waku-query-engine-test",
-  "version": "0.0.172",
+  "version": "0.0.174",
  "type": "module",
  "private": true,
  "scripts": {
@@ -10,7 +10,7 @@ import { mockLLMEvent } from "./utils.js";
 let llm: LLM;
 beforeEach(async () => {
  Settings.llm = new Anthropic({
-    model: "claude-3-opus",
+    model: "claude-3.5-sonnet",
  });
  llm = Settings.llm;
 });
@@ -1,5 +1,71 @@
 # examples

+## 0.3.27
+
+### Patch Changes
+
+- Updated dependencies [229cdeb]
+- Updated dependencies [47a7555]
+  - @llamaindex/groq@0.0.79
+  - @llamaindex/cloud@4.0.18
+  - llamaindex@0.11.13
+
+## 0.3.26
+
+### Patch Changes
+
+- Updated dependencies [d578889]
+- Updated dependencies [0fcc92f]
+- Updated dependencies [515a8b9]
+- Updated dependencies [3cd8a57]
+- Updated dependencies [f2dfd30]
+  - @llamaindex/core@0.6.13
+  - llamaindex@0.11.12
+  - @llamaindex/tools@0.1.3
+  - @llamaindex/bm25-retriever@0.0.2
+  - @llamaindex/cloud@4.0.17
+  - @llamaindex/node-parser@2.0.13
+  - @llamaindex/anthropic@0.3.15
+  - @llamaindex/assemblyai@0.1.12
+  - @llamaindex/clip@0.0.63
+  - @llamaindex/cohere@0.0.27
+  - @llamaindex/deepinfra@0.0.63
+  - @llamaindex/discord@0.1.12
+  - @llamaindex/google@0.3.12
+  - @llamaindex/huggingface@0.1.17
+  - @llamaindex/jinaai@0.0.23
+  - @llamaindex/mistral@0.1.13
+  - @llamaindex/mixedbread@0.0.27
+  - @llamaindex/notion@0.1.12
+  - @llamaindex/ollama@0.1.13
+  - @llamaindex/openai@0.4.7
+  - @llamaindex/perplexity@0.0.20
+  - @llamaindex/portkey-ai@0.0.55
+  - @llamaindex/replicate@0.0.55
+  - @llamaindex/astra@0.0.27
+  - @llamaindex/azure@0.1.24
+  - @llamaindex/chroma@0.0.27
+  - @llamaindex/elastic-search@0.1.13
+  - @llamaindex/firestore@1.0.20
+  - @llamaindex/milvus@0.1.22
+  - @llamaindex/mongodb@0.0.28
+  - @llamaindex/pinecone@0.1.13
+  - @llamaindex/postgres@0.0.56
+  - @llamaindex/qdrant@0.1.23
+  - @llamaindex/supabase@0.1.13
+  - @llamaindex/upstash@0.0.27
+  - @llamaindex/weaviate@0.0.28
+  - @llamaindex/vercel@0.1.13
+  - @llamaindex/voyage-ai@1.0.19
+  - @llamaindex/readers@3.1.12
+  - @llamaindex/workflow@1.1.13
+  - @llamaindex/deepseek@0.0.23
+  - @llamaindex/fireworks@0.0.23
+  - @llamaindex/groq@0.0.78
+  - @llamaindex/together@0.0.23
+  - @llamaindex/vllm@0.0.49
+  - @llamaindex/xai@0.0.10
+
 ## 0.3.25

 ### Patch Changes
@@ -0,0 +1,36 @@
+import { openai } from "@llamaindex/openai";
+import { agent } from "@llamaindex/workflow";
+import { createMemory, staticBlock } from "llamaindex";
+
+// Simple example: Agent with Predefined Memory
+async function simpleAgentMemoryExample() {
+  console.log("=== Simple Agent Memory Example ===");
+
+  const memory = createMemory({
+    memoryBlocks: [
+      staticBlock({
+        content:
+          "The user is a software engineer who loves TypeScript and LlamaIndex.",
+      }),
+    ],
+  });
+
+  // Create agent workflow
+  const workflow = agent({
+    name: "assistant",
+    llm: openai({ model: "gpt-4.1-nano" }),
+    memory,
+  });
+
+  // Test - agent should remember John and the shopping cart context
+  console.log("\n--- Testing Memory Context ---");
+  const result = await workflow.run("Hi, my name is John. Do you know me?");
+
+  console.log("Assistant Response:", result.data.result);
+
+  const result2 = await workflow.run("What is my name?");
+  console.log("Assistant Response:", result2.data.result);
+}
+
+// Run the example
+simpleAgentMemoryExample().catch(console.error);
@@ -0,0 +1,58 @@
+import { openai } from "@llamaindex/openai";
+import { createMemory } from "llamaindex";
+
+// Example: Basic Memory Usage with Factory
+async function basicMemoryExample() {
+  console.log("\n=== Example: Basic Memory Usage with Factory ===");
+
+  const memory = createMemory({ tokenLimit: 30 });
+
+  // Add messages to memory
+  await memory.add({
+    role: "user",
+    content: "Hi, my name is John and I'm a software engineer.",
+  });
+
+  await memory.add({
+    role: "assistant",
+    content: "Hello John! Nice to meet you. How can I help you today?",
+  });
+
+  await memory.add({
+    role: "user",
+    content: "I love working with TypeScript and React.",
+  });
+  // Not all messages are included because of token limit is set to 30
+  const llmMessages = await memory.getLLM();
+  console.log(
+    `\nLLM messages (${llmMessages.length} messages) limited by a small token limit:`,
+  );
+  llmMessages.forEach((msg, idx) => {
+    console.log(`${idx + 1}. ${msg.role}: ${msg.content}`);
+  });
+
+  // But the token limit above will be the window size of an LLM instance if you use getLLM with LLM
+  const llm = openai({ model: "gpt-4.1-mini" });
+  const llmMessagesWithLLM = await memory.getLLM(llm);
+  // Now all the messages are included because of the LLM window size of the model is much larger
+  console.log(
+    `\nLLM messages with LLM (${llmMessagesWithLLM.length} messages) limited by LLM window size:`,
+  );
+  llmMessagesWithLLM.forEach((msg, idx) => {
+    console.log(`${idx + 1}. ${msg.role}: ${msg.content}`);
+  });
+}
+
+// Main function
+async function main() {
+  console.log("🧠 Basic Memory Factory Examples");
+  console.log("===============================");
+
+  try {
+    await basicMemoryExample();
+  } catch (error) {
+    console.error("Error running basic memory examples:", error);
+  }
+}
+
+main().catch(console.error);
@@ -0,0 +1,101 @@
+import { openai } from "@llamaindex/openai";
+import { createMemory, factExtractionBlock } from "llamaindex";
+
+// Configure OpenAI
+const llm = openai({ model: "gpt-4.1-mini" });
+
+// Example: Memory with Fact Extraction
+async function factExtractionMemoryExample() {
+  console.log("\n=== Memory with Fact Extraction ===");
+
+  // Create memory with a fact extraction
+  const memory = createMemory([], {
+    tokenLimit: 100,
+    shortTermTokenLimitRatio: 0.7, // 70% for short-term, 30% for long-term
+    memoryBlocks: [
+      factExtractionBlock({
+        id: "user-facts",
+        priority: 5,
+        llm: llm,
+        maxFacts: 10,
+        isLongTerm: true,
+      }),
+    ],
+  });
+
+  // Simulate a conversation with facts
+  const conversationTurns = [
+    {
+      role: "user",
+      content: "Hi, I'm Sarah and I work as a data scientist at Google.",
+    },
+    {
+      role: "assistant",
+      content:
+        "Hello Sarah! It's great to meet you. Data science at Google must be exciting!",
+    },
+    {
+      role: "user",
+      content:
+        "Yes, I specialize in machine learning and natural language processing.",
+    },
+    {
+      role: "assistant",
+      content: "That's impressive! ML and NLP are fascinating fields.",
+    },
+    {
+      role: "user",
+      content:
+        "I have a PhD in Computer Science from Stanford, and I love hiking on weekends.",
+    },
+    {
+      role: "assistant",
+      content:
+        "Wow, Stanford PhD! And hiking is a great way to unwind from tech work.",
+    },
+    {
+      role: "user",
+      content: "I also have two cats named Whiskers and Mittens.",
+    },
+    {
+      role: "assistant",
+      content:
+        "Cats make wonderful companions! Whiskers and Mittens are cute names.",
+    },
+  ];
+
+  // Add conversation turns to memory
+  console.log("Adding conversation to memory...");
+  for (const turn of conversationTurns) {
+    await memory.add(turn);
+  }
+
+  // Get messages - facts should be extracted and included
+  const messages = await memory.getLLM(llm);
+  console.log("\nMessages with extracted facts:");
+  messages.forEach((msg, idx) => {
+    console.log(`${idx + 1}. ${msg.role ?? "unknown"}: ${msg.content}`);
+  });
+  //Messages with extracted facts:
+  // 1. assistant: Cats make wonderful companions! Whiskers and Mittens are cute names.
+  // 2. user: I also have two cats named Whiskers and Mittens.
+  // 3. assistant: Wow, Stanford PhD! And hiking is a great way to unwind from tech work.
+  // 4. memory: Sarah works as a data scientist at Google
+  // Sarah specializes in machine learning and natural language processing
+  // Sarah has a PhD in Computer Science from Stanford
+  // Sarah enjoys hiking on weekends
+}
+
+// Main function
+async function main() {
+  console.log("🧠 Fact Extraction Memory Example");
+  console.log("=================================");
+
+  try {
+    await factExtractionMemoryExample();
+  } catch (error) {
+    console.error("Error running fact extraction memory example:", error);
+  }
+}
+
+main().catch(console.error);
@@ -0,0 +1,62 @@
+import { openai } from "@llamaindex/openai";
+import { createMemory, staticBlock } from "llamaindex";
+
+// Configure OpenAI
+const llm = openai({ model: "gpt-4.1-mini" });
+
+// Example: Memory with Static Blocks
+async function staticMemoryBlockExample() {
+  console.log("\n=== Memory with Static Blocks ===");
+  console.log("- Memory always include static block");
+  console.log("- Memory cut off the messages within token limit\n");
+
+  // Create memory with a static block
+  const memory = createMemory([], {
+    tokenLimit: 30, // A small token limit which is not enough for the whole conversation below
+    memoryBlocks: [
+      staticBlock({
+        content:
+          "The user's name is John and he is a software engineer who loves TypeScript and LlamaIndex.",
+      }),
+    ],
+  });
+
+  // Add some messages to the memory
+  await memory.add({
+    role: "user",
+    content: "What do you know about me?",
+  });
+
+  await memory.add({
+    role: "assistant",
+    content:
+      "Based on our conversation, I know you're John, a software engineer who enjoys working with TypeScript and LlamaIndex!",
+  });
+
+  await memory.add({
+    role: "user",
+    content: "Which language does LlamaIndex support?",
+  });
+
+  // Get messages
+  // static block will always be included
+  // only the last message will be included because of token limit set above
+  const messages = await memory.getLLM(llm);
+  messages.forEach((msg, idx) => {
+    console.log(`${idx + 1}. ${msg.role}: ${msg.content}`);
+  });
+  // Messages with static block:
+  // 1. user: The user's name is John and he is a software engineer who loves TypeScript and LlamaIndex.
+  // 2. user: Which language does LlamaIndex support?
+}
+
+// Main function
+async function main() {
+  try {
+    await staticMemoryBlockExample();
+  } catch (error) {
+    console.error("Error running static memory blocks example:", error);
+  }
+}
+
+main().catch(console.error);
@@ -59,7 +59,7 @@ async function main() {

  const anthropic = new Anthropic({
    apiKey: process.env.ANTHROPIC_API_KEY,
-    model: "claude-3-opus",
+    model: "claude-3.5-sonnet",
  });

  // Create an ReActAgent with the function tools
@@ -61,7 +61,7 @@ async function main() {
  // Create an OpenAIAgent with the function tools
  const agent = new ReActAgent({
    llm: new Anthropic({
-      model: "claude-3-opus",
+      model: "claude-3.5-sonnet",
    }),
    tools: [functionTool, functionTool2],
  });
@@ -1,5 +1,5 @@
 import { Anthropic } from "@llamaindex/anthropic";
-import { ChatMemoryBuffer, SimpleChatEngine } from "llamaindex";
+import { createMemory, SimpleChatEngine } from "llamaindex";
 import { stdin as input, stdout as output } from "node:process";
 import readline from "node:readline/promises";

@@ -9,14 +9,12 @@ import readline from "node:readline/promises";
    model: "claude-3-7-sonnet",
  });
  // chatHistory will store all the messages in the conversation
-  const chatHistory = new ChatMemoryBuffer({
-    chatHistory: [
-      {
-        content: "You want to talk in rhymes.",
-        role: "system",
-      },
-    ],
-  });
+  const chatHistory = createMemory([
+    {
+      content: "You want to talk in rhymes.",
+      role: "system",
+    },
+  ]);
  const chatEngine = new SimpleChatEngine({
    llm,
    memory: chatHistory,
@@ -1,6 +1,6 @@
 {
  "name": "@llamaindex/examples",
-  "version": "0.3.25",
+  "version": "0.3.27",
  "private": true,
  "scripts": {
    "lint": "eslint .",
@@ -11,51 +11,52 @@
    "@azure/cosmos": "^4.1.1",
    "@azure/identity": "^4.4.1",
    "@azure/search-documents": "^12.1.0",
-    "@llamaindex/anthropic": "^0.3.14",
-    "@llamaindex/assemblyai": "^0.1.11",
-    "@llamaindex/astra": "^0.0.26",
-    "@llamaindex/azure": "^0.1.23",
-    "@llamaindex/chroma": "^0.0.26",
-    "@llamaindex/clip": "^0.0.62",
-    "@llamaindex/cloud": "^4.0.16",
-    "@llamaindex/cohere": "^0.0.26",
-    "@llamaindex/core": "^0.6.12",
-    "@llamaindex/deepinfra": "^0.0.62",
-    "@llamaindex/deepseek": "^0.0.22",
-    "@llamaindex/discord": "^0.1.11",
-    "@llamaindex/elastic-search": "^0.1.12",
+    "@llamaindex/anthropic": "^0.3.15",
+    "@llamaindex/assemblyai": "^0.1.12",
+    "@llamaindex/astra": "^0.0.27",
+    "@llamaindex/azure": "^0.1.24",
+    "@llamaindex/bm25-retriever": "^0.0.2",
+    "@llamaindex/chroma": "^0.0.27",
+    "@llamaindex/clip": "^0.0.63",
+    "@llamaindex/cloud": "^4.0.18",
+    "@llamaindex/cohere": "^0.0.27",
+    "@llamaindex/core": "^0.6.13",
+    "@llamaindex/deepinfra": "^0.0.63",
+    "@llamaindex/deepseek": "^0.0.23",
+    "@llamaindex/discord": "^0.1.12",
+    "@llamaindex/elastic-search": "^0.1.13",
    "@llamaindex/env": "^0.1.30",
-    "@llamaindex/firestore": "^1.0.19",
-    "@llamaindex/fireworks": "^0.0.22",
-    "@llamaindex/google": "^0.3.11",
-    "@llamaindex/groq": "^0.0.77",
-    "@llamaindex/huggingface": "^0.1.16",
-    "@llamaindex/jinaai": "^0.0.22",
-    "@llamaindex/milvus": "^0.1.21",
-    "@llamaindex/mistral": "^0.1.12",
-    "@llamaindex/mixedbread": "^0.0.26",
-    "@llamaindex/mongodb": "^0.0.27",
-    "@llamaindex/node-parser": "^2.0.12",
-    "@llamaindex/notion": "^0.1.11",
-    "@llamaindex/ollama": "^0.1.12",
-    "@llamaindex/openai": "^0.4.6",
-    "@llamaindex/perplexity": "^0.0.19",
-    "@llamaindex/pinecone": "^0.1.12",
-    "@llamaindex/portkey-ai": "^0.0.54",
-    "@llamaindex/postgres": "^0.0.55",
-    "@llamaindex/qdrant": "^0.1.22",
-    "@llamaindex/readers": "^3.1.11",
-    "@llamaindex/replicate": "^0.0.54",
-    "@llamaindex/supabase": "^0.1.12",
-    "@llamaindex/together": "^0.0.22",
-    "@llamaindex/tools": "^0.1.1",
-    "@llamaindex/upstash": "^0.0.26",
-    "@llamaindex/vercel": "^0.1.12",
-    "@llamaindex/vllm": "^0.0.48",
-    "@llamaindex/voyage-ai": "^1.0.18",
-    "@llamaindex/weaviate": "^0.0.27",
-    "@llamaindex/workflow": "^1.1.12",
-    "@llamaindex/xai": "workspace:^0.0.9",
+    "@llamaindex/firestore": "^1.0.20",
+    "@llamaindex/fireworks": "^0.0.23",
+    "@llamaindex/google": "^0.3.12",
+    "@llamaindex/groq": "^0.0.79",
+    "@llamaindex/huggingface": "^0.1.17",
+    "@llamaindex/jinaai": "^0.0.23",
+    "@llamaindex/milvus": "^0.1.22",
+    "@llamaindex/mistral": "^0.1.13",
+    "@llamaindex/mixedbread": "^0.0.27",
+    "@llamaindex/mongodb": "^0.0.28",
+    "@llamaindex/node-parser": "^2.0.13",
+    "@llamaindex/notion": "^0.1.12",
+    "@llamaindex/ollama": "^0.1.13",
+    "@llamaindex/openai": "^0.4.7",
+    "@llamaindex/perplexity": "^0.0.20",
+    "@llamaindex/pinecone": "^0.1.13",
+    "@llamaindex/portkey-ai": "^0.0.55",
+    "@llamaindex/postgres": "^0.0.56",
+    "@llamaindex/qdrant": "^0.1.23",
+    "@llamaindex/readers": "^3.1.12",
+    "@llamaindex/replicate": "^0.0.55",
+    "@llamaindex/supabase": "^0.1.13",
+    "@llamaindex/together": "^0.0.23",
+    "@llamaindex/tools": "^0.1.3",
+    "@llamaindex/upstash": "^0.0.27",
+    "@llamaindex/vercel": "^0.1.13",
+    "@llamaindex/vllm": "^0.0.49",
+    "@llamaindex/voyage-ai": "^1.0.19",
+    "@llamaindex/weaviate": "^0.0.28",
+    "@llamaindex/workflow": "^1.1.13",
+    "@llamaindex/xai": "workspace:^0.0.10",
    "@notionhq/client": "^2.2.15",
    "@pinecone-database/pinecone": "^4.0.0",
    "@vercel/postgres": "^0.10.0",
@@ -64,7 +65,7 @@
    "commander": "^12.1.0",
    "dotenv": "^16.4.5",
    "js-tiktoken": "^1.0.14",
-    "llamaindex": "^0.11.11",
+    "llamaindex": "^0.11.13",
    "mongodb": "6.7.0",
    "postgres": "^3.4.4",
    "wikipedia": "^2.1.2",
@@ -2,11 +2,7 @@ import { stdin as input, stdout as output } from "node:process";
 import readline from "node:readline/promises";

 import { OpenAI } from "@llamaindex/openai";
-import {
-  ChatSummaryMemoryBuffer,
-  Settings,
-  SimpleChatEngine,
-} from "llamaindex";
+import { createMemory, Settings, SimpleChatEngine } from "llamaindex";

 if (process.env.NODE_ENV === "development") {
  Settings.callbackManager.on("llm-end", (event) => {
@@ -15,10 +11,13 @@ if (process.env.NODE_ENV === "development") {
 }

 async function main() {
-  // Set maxTokens to 75% of the context window size of 4096
-  // This will trigger the summarizer once the chat history reaches 25% of the context window size (1024 tokens)
-  const llm = new OpenAI({ model: "gpt-3.5-turbo", maxTokens: 4096 * 0.75 });
-  const chatHistory = new ChatSummaryMemoryBuffer({ llm });
+  const llm = new OpenAI({ model: "gpt-3.5-turbo" });
+  const chatHistory = createMemory([
+    {
+      content: "You are a helpful assistant.",
+      role: "system",
+    },
+  ]);
  const chatEngine = new SimpleChatEngine({ llm });
  const rl = readline.createInterface({ input, output });

@@ -29,10 +28,6 @@ async function main() {
      chatHistory,
      stream: true,
    });
-    if (chatHistory.getLastSummary()) {
-      // Print the summary of the conversation so far that is produced by the SummaryChatHistory
-      console.log(`Summary: ${chatHistory.getLastSummary()?.content}`);
-    }
    for await (const chunk of stream) {
      process.stdout.write(chunk.response);
    }
@@ -1,4 +1,4 @@
-import { LlamaParseReader } from "@llamaindex/cloud";
+import { LlamaParseReader } from "@llamaindex/cloud/reader";
 import { openai, OpenAIEmbedding } from "@llamaindex/openai";
 import { Settings, VectorStoreIndex } from "llamaindex";

@@ -1,4 +1,4 @@
-import { LlamaParseReader } from "@llamaindex/cloud";
+import { LlamaParseReader } from "@llamaindex/cloud/reader";
 import { SimpleDirectoryReader } from "@llamaindex/readers/directory";
 import { VectorStoreIndex } from "llamaindex";

@@ -0,0 +1,14 @@
+# BM25 Retriever
+
+In this guide, we introduce a bm25 retriever that search documents using the bm25 method. BM25 (Best Matching 25) is a ranking function that extends TF-IDF by considering term frequency saturation and document length. BM25 effectively ranks documents based on query term occurrence and rarity across the corpus.
+
+## Setup
+
+1. `cd` Into the `examples` directory
+2. run `npm i`
+
+## Example
+
+```bash
+`npx tsx ./retrievers/bm25/example.ts`
+```
@@ -0,0 +1,33 @@
+import { Bm25Retriever } from "@llamaindex/bm25-retriever";
+import { OpenAIEmbedding } from "@llamaindex/openai";
+import { PDFReader } from "@llamaindex/readers/pdf";
+import { MetadataMode, Settings, VectorStoreIndex } from "llamaindex";
+
+Settings.embedModel = new OpenAIEmbedding();
+
+async function main() {
+  // Load PDF
+  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);
+
+  const retriever = new Bm25Retriever({
+    docStore: index.docStore,
+    topK: 3,
+  });
+
+  // Query the data
+  const response = await retriever.retrieve({
+    query: "What mistakes did Warren E. Buffett make?",
+  });
+
+  // Output response
+  response.forEach((r) => {
+    console.log(`Score: ${r.score}`);
+    console.log(`Text: ${r.node.getContent(MetadataMode.NONE)}`);
+  });
+}
+
+main().catch(console.error);
@@ -1,5 +1,18 @@
 # @llamaindex/autotool

+## 8.0.13
+
+### Patch Changes
+
+- llamaindex@0.11.13
+
+## 8.0.12
+
+### Patch Changes
+
+- Updated dependencies [515a8b9]
+  - llamaindex@0.11.12
+
 ## 8.0.11

 ### Patch Changes
@@ -1,5 +1,20 @@
 # @llamaindex/autotool-01-node-example

+## 0.0.121
+
+### Patch Changes
+
+- llamaindex@0.11.13
+- @llamaindex/autotool@8.0.13
+
+## 0.0.120
+
+### Patch Changes
+
+- Updated dependencies [515a8b9]
+  - llamaindex@0.11.12
+  - @llamaindex/autotool@8.0.12
+
 ## 0.0.119

 ### Patch Changes
@@ -13,5 +13,5 @@
  "scripts": {
    "start": "node --import tsx --import @llamaindex/autotool/node ./src/index.ts"
  },
-  "version": "0.0.119"
+  "version": "0.0.121"
 }
@@ -6,7 +6,7 @@
    "url": "git+https://github.com/run-llama/LlamaIndexTS.git",
    "directory": "packages/autotool"
  },
-  "version": "8.0.11",
+  "version": "8.0.13",
  "description": "auto transpile your JS function to LLM Agent compatible",
  "files": [
    "dist",
@@ -1,5 +1,20 @@
 # @llamaindex/cloud

+## 4.0.18
+
+### Patch Changes
+
+- 47a7555: chore: bump sdk version
+
+## 4.0.17
+
+### Patch Changes
+
+- Updated dependencies [d578889]
+- Updated dependencies [0fcc92f]
+- Updated dependencies [515a8b9]
+  - @llamaindex/core@0.6.13
+
 ## 4.0.16

 ### Patch Changes
@@ -1,6 +1,6 @@
 {
  "name": "@llamaindex/cloud",
-  "version": "4.0.16",
+  "version": "4.0.18",
  "type": "module",
  "license": "MIT",
  "scripts": {
@@ -171,6 +171,14 @@ export class LlamaParseReader extends FileReader {
  replace_failed_page_with_error_message_suffix?: string | undefined;
  save_images?: boolean | undefined;
  preset?: string | undefined;
+  high_res_ocr?: boolean | undefined;
+  outlined_table_extraction?: boolean | undefined;
+  hide_headers?: boolean | undefined;
+  hide_footers?: boolean | undefined;
+  page_header_prefix?: string | undefined;
+  page_header_suffix?: string | undefined;
+  page_footer_prefix?: string | undefined;
+  page_footer_suffix?: string | undefined;

  constructor(
    params: Partial<Omit<LlamaParseReader, "language" | "apiKey">> & {
@@ -352,6 +360,14 @@ export class LlamaParseReader extends FileReader {
        this.replace_failed_page_with_error_message_suffix,
      save_images: this.save_images,
      preset: this.preset,
+      high_res_ocr: this.high_res_ocr,
+      outlined_table_extraction: this.outlined_table_extraction,
+      hide_headers: this.hide_headers,
+      hide_footers: this.hide_footers,
+      page_header_prefix: this.page_header_prefix,
+      page_header_suffix: this.page_header_suffix,
+      page_footer_prefix: this.page_footer_prefix,
+      page_footer_suffix: this.page_footer_suffix,
    } satisfies {
      [Key in keyof BodyUploadFileApiParsingUploadPost]-?:
        | BodyUploadFileApiParsingUploadPost[Key]
@@ -1,5 +1,13 @@
 # @llamaindex/core

+## 0.6.13
+
+### Patch Changes
+
+- d578889: Add new memory API
+- 0fcc92f: Fix: split sentences must not trim whitespaces
+- 515a8b9: Fix: logging for fromPersistPath
+
 ## 0.6.12

 ### Patch Changes
@@ -426,7 +434,6 @@
 ### Patch Changes

 - 8b7fdba: refactor: move chat engine & retriever into core.
-
  - `chatHistory` in BaseChatEngine now returns `ChatMessage[] | Promise<ChatMessage[]>`, instead of `BaseMemory`
  - update `retrieve-end` type

@@ -452,7 +459,6 @@
 ### Patch Changes

 - 2cd1383: refactor: align `response-synthesizers` & `chat-engine` module
-
  - builtin event system
  - correct class extends
  - aligin APIs, naming with llama-index python
@@ -1,7 +1,7 @@
 {
  "name": "@llamaindex/core",
  "type": "module",
-  "version": "0.6.12",
+  "version": "0.6.13",
  "description": "LlamaIndex Core Module",
  "exports": {
    "./agent": {
@@ -152,6 +152,7 @@ export type AgentParamsBase<

 /**
 * Worker will schedule tasks and handle the task execution
+ * @deprecated Use agent instead.
 */
 export abstract class AgentWorker<
  AI extends LLM,
@@ -250,6 +251,7 @@ export abstract class AgentWorker<

 /**
 * Runner will manage the task execution and provide a high-level API for the user
+ * @deprecated Use agent instead.
 */
 export abstract class AgentRunner<
  AI extends LLM,
@@ -62,6 +62,9 @@ export class LLMAgentWorker extends AgentWorker<LLM> {
  taskHandler = AgentRunner.defaultTaskHandler;
 }

+/**
+ * @deprecated Use agent instead.
+ */
 export class LLMAgent extends AgentRunner<LLM> {
  constructor(params: LLMAgentParams<LLM>) {
    validateAgentParams(params);
@@ -1,5 +1,5 @@
 import type { ChatMessage, MessageContent } from "../llms";
-import type { BaseMemory } from "../memory";
+import type { Memory } from "../memory";
 import { EngineResponse } from "../schema";

 export interface BaseChatEngineParams<
@@ -9,9 +9,7 @@ export interface BaseChatEngineParams<
  /**
   * Optional chat history if you want to customize the chat history.
   */
-  chatHistory?:
-    | ChatMessage<AdditionalMessageOptions>[]
-    | BaseMemory<AdditionalMessageOptions>;
+  chatHistory?: ChatMessage<AdditionalMessageOptions>[] | Memory;
 }

 export interface StreamingChatEngineParams<
@@ -1,7 +1,7 @@
 import { wrapEventCaller } from "../decorator";
 import { Settings } from "../global";
 import type { ChatMessage, LLM, MessageContent, MessageType } from "../llms";
-import { BaseMemory, ChatMemoryBuffer } from "../memory";
+import { Memory, createMemory } from "../memory";
 import type { BaseNodePostprocessor } from "../postprocessor";
 import {
  type ContextSystemPrompt,
@@ -23,7 +23,7 @@ import type { ContextGenerator } from "./type";
 export type ContextChatEngineOptions = {
  retriever: BaseRetriever;
  chatModel?: LLM | undefined;
-  chatHistory?: ChatMessage[] | undefined;
+  chatHistory?: ChatMessage[] | Memory | undefined;
  contextSystemPrompt?: ContextSystemPrompt | undefined;
  nodePostprocessors?: BaseNodePostprocessor[] | undefined;
  systemPrompt?: string | undefined;
@@ -37,18 +37,21 @@ export type ContextChatEngineOptions = {
 */
 export class ContextChatEngine extends PromptMixin implements BaseChatEngine {
  chatModel: LLM;
-  memory: BaseMemory;
+  memory: Memory;
  contextGenerator: ContextGenerator & PromptMixin;
  systemPrompt?: string | undefined;

  get chatHistory() {
-    return this.memory.getMessages();
+    return this.memory.getLLM();
  }

  constructor(init: ContextChatEngineOptions) {
    super();
    this.chatModel = init.chatModel ?? Settings.llm;
-    this.memory = new ChatMemoryBuffer({ chatHistory: init?.chatHistory });
+    this.memory =
+      init?.chatHistory instanceof Memory
+        ? init.chatHistory
+        : createMemory(init?.chatHistory ?? []);
    this.contextGenerator = new DefaultContextGenerator({
      retriever: init.retriever,
      contextSystemPrompt: init?.contextSystemPrompt,
@@ -87,12 +90,9 @@ export class ContextChatEngine extends PromptMixin implements BaseChatEngine {
  ): Promise<EngineResponse | AsyncIterable<EngineResponse>> {
    const { message, stream } = params;
    const chatHistory = params.chatHistory
-      ? new ChatMemoryBuffer({
-          chatHistory:
-            params.chatHistory instanceof BaseMemory
-              ? await params.chatHistory.getMessages()
-              : params.chatHistory,
-        })
+      ? params.chatHistory instanceof Memory
+        ? params.chatHistory
+        : createMemory(params.chatHistory)
      : this.memory;
    const requestMessages = await this.prepareRequestMessages(
      message,
@@ -110,7 +110,7 @@ export class ContextChatEngine extends PromptMixin implements BaseChatEngine {
          initialValue: "",
          reducer: (accumulator, part) => (accumulator += part.delta),
          finished: (accumulator) => {
-            chatHistory.put({ content: accumulator, role: "assistant" });
+            void chatHistory.add({ content: accumulator, role: "assistant" });
          },
        }),
        (r) => EngineResponse.fromChatResponseChunk(r, requestMessages.nodes),
@@ -120,26 +120,26 @@ export class ContextChatEngine extends PromptMixin implements BaseChatEngine {
      messages: requestMessages.messages,
      additionalChatOptions: params.chatOptions as object,
    });
-    chatHistory.put(response.message);
+    await chatHistory.add(response.message);
    return EngineResponse.fromChatResponse(response, requestMessages.nodes);
  }

-  reset() {
-    this.memory.reset();
+  async reset() {
+    await this.memory.clear();
  }

  private async prepareRequestMessages(
    message: MessageContent,
-    chatHistory: BaseMemory,
+    chatHistory: Memory,
  ) {
-    chatHistory.put({
+    await chatHistory.add({
      content: message,
      role: "user",
    });
    const textOnly = extractText(message);
    const context = await this.contextGenerator.generate(textOnly);
    const systemMessage = this.prependSystemPrompt(context.message);
-    const messages = await chatHistory.getMessages([systemMessage]);
+    const messages = await chatHistory.getLLM(this.chatModel, [systemMessage]);
    return { nodes: context.nodes, messages };
  }

@@ -1,5 +1,5 @@
 import type { LLM } from "../llms";
-import { BaseMemory, ChatMemoryBuffer } from "../memory";
+import { createMemory, Memory } from "../memory";
 import { EngineResponse } from "../schema";
 import { streamConverter, streamReducer } from "../utils";
 import type {
@@ -16,20 +16,16 @@ import { Settings } from "../global";
 */

 export class SimpleChatEngine implements BaseChatEngine {
-  memory: BaseMemory;
+  memory: Memory;
  llm: LLM;

  get chatHistory() {
-    return this.memory.getMessages();
+    return this.memory.getLLM();
  }

  constructor(init?: Partial<SimpleChatEngine>) {
    this.llm = init?.llm ?? Settings.llm;
-    this.memory =
-      init?.memory ??
-      new ChatMemoryBuffer({
-        llm: this.llm,
-      });
+    this.memory = init?.memory ?? createMemory();
  }

  chat(params: NonStreamingChatEngineParams): Promise<EngineResponse>;
@@ -43,19 +39,15 @@ export class SimpleChatEngine implements BaseChatEngine {
    const { message, stream } = params;

    const chatHistory = params.chatHistory
-      ? new ChatMemoryBuffer({
-          llm: this.llm,
-          chatHistory:
-            params.chatHistory instanceof BaseMemory
-              ? await params.chatHistory.getMessages()
-              : params.chatHistory,
-        })
+      ? params.chatHistory instanceof Memory
+        ? params.chatHistory
+        : createMemory(params.chatHistory)
      : this.memory;
-    chatHistory.put({ content: message, role: "user" });
+    await chatHistory.add({ content: message, role: "user" });

    if (stream) {
      const stream = await this.llm.chat({
-        messages: await chatHistory.getMessages(),
+        messages: await chatHistory.getLLM(this.llm),
        stream: true,
      });
      return streamConverter(
@@ -64,7 +56,7 @@ export class SimpleChatEngine implements BaseChatEngine {
          initialValue: "",
          reducer: (accumulator, part) => accumulator + part.delta,
          finished: (accumulator) => {
-            chatHistory.put({ content: accumulator, role: "assistant" });
+            void chatHistory.add({ content: accumulator, role: "assistant" });
          },
        }),
        EngineResponse.fromChatResponseChunk,
@@ -73,13 +65,13 @@ export class SimpleChatEngine implements BaseChatEngine {

    const response = await this.llm.chat({
      stream: false,
-      messages: await chatHistory.getMessages(),
+      messages: await chatHistory.getLLM(this.llm),
    });
-    chatHistory.put(response.message);
+    await chatHistory.add(response.message);
    return EngineResponse.fromChatResponse(response);
  }

-  reset() {
-    this.memory.reset();
+  async reset() {
+    await this.memory.clear();
  }
 }
@@ -0,0 +1,7 @@
+import type { MemoryMessage } from "../types";
+
+export interface MessageAdapter<T, TMessageOptions extends object = object> {
+  fromMemory(message: MemoryMessage<TMessageOptions>): T;
+  toMemory(message: T): MemoryMessage<TMessageOptions>;
+  isCompatible(message: unknown): message is T;
+}
@@ -0,0 +1,43 @@
+import { randomUUID } from "@llamaindex/env";
+import type { ChatMessage } from "../../llms";
+import type { MemoryMessage } from "../types";
+import { type MessageAdapter } from "./base";
+
+export class ChatMessageAdapter<
+  AdditionalMessageOptions extends object = object,
+> implements
+    MessageAdapter<
+      ChatMessage<AdditionalMessageOptions>,
+      AdditionalMessageOptions
+    >
+{
+  fromMemory(
+    message: MemoryMessage<AdditionalMessageOptions>,
+  ): ChatMessage<AdditionalMessageOptions> {
+    return {
+      content: message.content,
+      role: message.role,
+      options: message.options,
+    };
+  }
+  toMemory(
+    message: ChatMessage<AdditionalMessageOptions>,
+  ): MemoryMessage<AdditionalMessageOptions> {
+    return {
+      id: randomUUID(),
+      createdAt: new Date(),
+      ...message,
+    };
+  }
+  isCompatible(
+    message: unknown,
+  ): message is ChatMessage<AdditionalMessageOptions> {
+    return !!(
+      message &&
+      typeof message === "object" &&
+      "role" in message &&
+      message.role &&
+      "content" in message
+    );
+  }
+}
@@ -0,0 +1,3 @@
+export * from "./base";
+export * from "./chat";
+export * from "./vercel";
@@ -0,0 +1,198 @@
+import type {
+  ChatMessage,
+  MessageContent,
+  MessageContentDetail,
+} from "../../llms";
+import { extractText } from "../../utils";
+import type { MemoryMessage } from "../types";
+import type { MessageAdapter } from "./base";
+
+// UIMessage from the vercel/ai package (external)
+export type VercelMessage = {
+  id: string;
+  role: "system" | "user" | "assistant" | "data";
+  content: string;
+  createdAt?: Date | undefined;
+  annotations?: Array<unknown> | undefined;
+  parts: Array<{ type: string; [key: string]: unknown }>;
+};
+
+/**
+ * Utility class for converting between LlamaIndex ChatMessage and Vercel UI Message formats
+ */
+export class VercelMessageAdapter<
+  AdditionalMessageOptions extends object = object,
+> implements MessageAdapter<VercelMessage, AdditionalMessageOptions>
+{
+  /**
+   * Convert LlamaIndex ChatMessage to Vercel UI Message format
+   */
+  fromMemory(memoryMessage: MemoryMessage<object>): VercelMessage {
+    const parts = this.convertMessageContentToVercelParts(
+      memoryMessage.content,
+    );
+
+    // Convert role to UI message role
+    let role: VercelMessage["role"];
+    switch (memoryMessage.role) {
+      case "system":
+      case "user":
+      case "assistant":
+        role = memoryMessage.role;
+        break;
+      case "memory":
+        role = "system";
+        break;
+      case "developer":
+        role = "user";
+        break;
+      default:
+        role = "user"; // Default fallback, should not happen
+    }
+
+    return {
+      id: memoryMessage.id,
+      role,
+      content: extractText(memoryMessage.content),
+      parts,
+      createdAt: memoryMessage.createdAt,
+      annotations: memoryMessage.annotations,
+    };
+  }
+  /**
+   * Convert Vercel UI Message to LlamaIndex ChatMessage format
+   */
+  toMemory(uiMessage: VercelMessage): MemoryMessage<AdditionalMessageOptions> {
+    // Convert UI message role to MessageType
+    let role: ChatMessage["role"];
+    switch (uiMessage.role) {
+      case "system":
+      case "user":
+      case "assistant":
+        role = uiMessage.role;
+        break;
+      case "data":
+        role = "user"; // Map data role to user
+        break;
+      default:
+        role = "user"; // Default fallback, should not happen
+    }
+
+    // Convert parts to MessageContent
+    const content = this.convertVercelPartsToMessageContent(uiMessage.parts);
+
+    return {
+      id: uiMessage.id,
+      content: content ?? uiMessage.content,
+      role,
+      createdAt: uiMessage.createdAt,
+      annotations: uiMessage.annotations,
+    };
+  }
+
+  /**
+   * Validate if object matches VercelMessage structure
+   */
+  isCompatible(message: unknown): message is VercelMessage {
+    return !!(
+      message &&
+      typeof message === "object" &&
+      "role" in message &&
+      "content" in message &&
+      "parts" in message
+    );
+  }
+
+  /**
+   * Convert UI parts to MessageContent
+   */
+  private convertVercelPartsToMessageContent(
+    parts: VercelMessage["parts"],
+  ): MessageContent | null {
+    if (parts.length === 0) {
+      return null;
+    }
+
+    const details: MessageContentDetail[] = [];
+
+    for (const part of parts) {
+      switch (part.type) {
+        case "file": {
+          details.push({
+            type: "file",
+            data: part.data as string,
+            mimeType: part.mimeType as string,
+          });
+          break;
+        }
+        default:
+          // For other part types, convert to text
+          details.push({
+            type: "text",
+            text: part.text as string,
+          });
+          break;
+      }
+    }
+
+    // If only one text detail, return as string
+    if (details.length === 1 && details[0]?.type === "text") {
+      return details[0].text;
+    }
+
+    return details;
+  }
+
+  /**
+   * Convert MessageContent to UI parts
+   */
+  private convertMessageContentToVercelParts(
+    content: MessageContent,
+  ): VercelMessage["parts"] {
+    if (typeof content === "string") {
+      return [
+        {
+          type: "text",
+          text: content,
+        },
+      ];
+    }
+
+    const parts: VercelMessage["parts"] = [];
+
+    for (const detail of content) {
+      switch (detail.type) {
+        case "text":
+          parts.push({
+            type: "text",
+            text: detail.text,
+          });
+          break;
+        case "image_url":
+          parts.push({
+            type: "text",
+            text: `[Image URL: ${detail.image_url.url}]`,
+          });
+          break;
+        case "audio":
+        case "video":
+        case "image":
+        case "file":
+          parts.push({
+            type: "file",
+            data: detail.data,
+            mimeType: detail.type,
+          });
+          break;
+        default:
+          // For unknown types, create a text representation
+          parts.push({
+            type: "text",
+            text: JSON.stringify(detail),
+          });
+      }
+    }
+
+    return parts;
+  }
+}
@@ -0,0 +1,50 @@
+import { randomUUID } from "@llamaindex/env";
+import type { MemoryMessage } from "../types";
+
+export type MemoryBlockOptions = {
+  /**
+   * The id of the memory block.
+   */
+  id?: string;
+  /**
+   * The priority of the memory block.
+   * Note: if priority is 0, the block content is always included in the memory context.
+   */
+  priority: number;
+  /**
+   * Whether the memory block is long term.
+   * Default is true.
+   */
+  isLongTerm?: boolean;
+};
+
+/**
+ * A base class for memory blocks.
+ */
+export abstract class BaseMemoryBlock<
+  TAdditionalMessageOptions extends object = object,
+> {
+  public readonly id: string;
+  public readonly priority: number;
+  public readonly isLongTerm: boolean;
+
+  constructor(options: MemoryBlockOptions) {
+    this.id = options.id ?? `memory-block-${randomUUID()}`;
+    this.priority = options.priority;
+    this.isLongTerm = options.isLongTerm ?? true;
+  }
+
+  /**
+   * Pull the memory block content (async).
+   *
+   * @returns The memory block content as an array of ChatMessage.
+   */
+  abstract get(): Promise<MemoryMessage<TAdditionalMessageOptions>[]>;
+
+  /**
+   * Store the messages in the memory block.
+   */
+  abstract put(
+    messages: MemoryMessage<TAdditionalMessageOptions>[],
+  ): Promise<void>;
+}
@@ -0,0 +1,153 @@
+import type { LLM, MessageType } from "../../llms";
+import type { MemoryMessage } from "../types";
+import { BaseMemoryBlock, type MemoryBlockOptions } from "./base";
+
+const DEFAULT_EXTRACTION_PROMPT = `
+You are a precise fact extraction system designed to identify key information from conversations.
+
+CONVERSATION SEGMENT:
+{{conversation}}
+
+EXISTING FACTS:
+{{existing_facts}}
+
+INSTRUCTIONS: 
+1. Review the conversation segment provided above.
+2. Extract specific, concrete facts the user has disclosed or important information discovered
+3. Focus on factual information like preferences, personal details, requirements, constraints, or context
+4. Do not include opinions, summaries, or interpretations - only extract explicit information
+5. Do not duplicate facts that are already in the existing facts list
+
+Respond with the new facts from the conversation segment using the following JSON format:
+{
+  "facts": ["fact1", "fact2", "fact3", ...]
+}
+`;
+
+const DEFAULT_SUMMARY_PROMPT = `
+You are a precise fact condensing system designed to summarize facts in a concise manner.
+
+EXISTING FACTS:
+{{existing_facts}}
+
+INSTRUCTIONS:
+1. Review the current list of existing facts
+2. Condense the facts into a more concise list, less than {{ max_facts }} facts
+3. Focus on factual information like preferences, personal details, requirements, constraints, or context
+4. Do not include opinions, summaries, or interpretations - only extract explicit information
+5. Do not duplicate facts that are already in the existing facts list
+
+Respond with the condensed facts using the following JSON format:
+{
+  "facts": ["fact1", "fact2", "fact3", ...]
+}
+`;
+
+/**
+ * The options for the fact extraction memory block.
+ */
+export type FactExtractionMemoryBlockOptions = {
+  /**
+   * The fact extraction model to use.
+   */
+  llm: LLM;
+  /**
+   * The maximum number of facts to extract.
+   */
+  maxFacts: number;
+  /**
+   * The prompt to use for fact extraction.
+   */
+  extractionPrompt?: string;
+  /**
+   * The prompt to use for fact summary.
+   */
+  summaryPrompt?: string;
+} & MemoryBlockOptions & {
+    isLongTerm?: true;
+  };
+
+/**
+ * A memory block that stores facts extracted from conversations.
+ */
+export class FactExtractionMemoryBlock<
+  TAdditionalMessageOptions extends object = object,
+> extends BaseMemoryBlock<TAdditionalMessageOptions> {
+  private readonly llm: LLM;
+  private facts: string[] = [];
+  private readonly maxFacts: number;
+  private readonly extractionPrompt: string;
+  private readonly summaryPrompt: string;
+
+  constructor(options: FactExtractionMemoryBlockOptions) {
+    super(options);
+    this.llm = options.llm;
+    this.maxFacts = options.maxFacts;
+    this.extractionPrompt =
+      options.extractionPrompt ?? DEFAULT_EXTRACTION_PROMPT;
+    this.summaryPrompt = options.summaryPrompt ?? DEFAULT_SUMMARY_PROMPT;
+  }
+
+  async get(): Promise<MemoryMessage<TAdditionalMessageOptions>[]> {
+    const fact = {
+      id: this.id,
+      content: this.facts.join("\n"),
+      role: "memory" as MessageType,
+    };
+    return [fact];
+  }
+
+  async put(
+    messages: MemoryMessage<TAdditionalMessageOptions>[],
+  ): Promise<void> {
+    if (messages.length === 0) {
+      return;
+    }
+    // Format existing facts
+    const existingFactsStr = `{ facts: [${this.facts.join(", ")}] }`;
+    // Format conversation
+    const conversation = `\n\t${messages.map((m) => m.content).join("\n\t")}`;
+    // Format prompt
+    const prompt = this.extractionPrompt
+      .replace("{{conversation}}", conversation)
+      .replace("{{existing_facts}}", existingFactsStr);
+    // Call the LLM
+    const response = await this.llm.complete({
+      prompt,
+    });
+    // Parse and validate the response
+    const newFacts = JSON.parse(response.text);
+    if (newFacts.facts === undefined || !Array.isArray(newFacts.facts)) {
+      throw new Error(
+        `[FactExtraction] Invalid response from LLM: ${response.text}`,
+      );
+    }
+    // No new facts, so no need to update the facts
+    if (newFacts.facts.length === 0) {
+      return;
+    }
+    // Update the facts
+    this.facts.push(...newFacts.facts);
+
+    // Condense the facts
+    if (this.facts.length > this.maxFacts) {
+      const existingFactsStr = `{ facts: [${this.facts.join(", ")}] }`;
+      const prompt = this.summaryPrompt
+        .replace("{{existing_facts}}", existingFactsStr)
+        .replace("{{max_facts}}", this.maxFacts.toString());
+      const response = await this.llm.complete({
+        prompt,
+      });
+      const condensedFacts = JSON.parse(response.text);
+      if (
+        condensedFacts.facts === undefined ||
+        !Array.isArray(condensedFacts.facts) ||
+        condensedFacts.facts.length === 0
+      ) {
+        throw new Error("Invalid response from LLM");
+      }
+      // Only get the first maxFacts facts (in case the LLM returned more)
+      this.facts = condensedFacts.facts.slice(0, this.maxFacts);
+    }
+  }
+}
@@ -0,0 +1,3 @@
+export { BaseMemoryBlock } from "./base";
+export { FactExtractionMemoryBlock } from "./fact";
+export { StaticMemoryBlock } from "./static";
@@ -0,0 +1,51 @@
+import type { MessageContent, MessageType } from "../../llms";
+import type { MemoryMessage } from "../types";
+import { BaseMemoryBlock, type MemoryBlockOptions } from "./base";
+
+export type StaticMemoryBlockOptions = {
+  /**
+   * The static content to store.
+   */
+  content: MessageContent;
+  /**
+   * The role of the message.
+   */
+  messageRole?: MessageType;
+} & Omit<MemoryBlockOptions, "priority" | "isLongTerm">;
+
+/**
+ * A memory block that stores static content that doesn't change.
+ * Static content is always included in the memory context.
+ */
+export class StaticMemoryBlock<
+  TAdditionalMessageOptions extends object = object,
+> extends BaseMemoryBlock<TAdditionalMessageOptions> {
+  private readonly content: MessageContent;
+  private readonly messageRole: MessageType;
+
+  constructor(options: StaticMemoryBlockOptions) {
+    super({ ...options, priority: 0, isLongTerm: false });
+    this.content = options.content;
+    this.messageRole = options.messageRole ?? "user";
+  }
+
+  /**
+   * Returns the static content.
+   * The messages parameter is ignored since this block contains static content.
+   */
+  async get(): Promise<MemoryMessage<TAdditionalMessageOptions>[]> {
+    return [
+      {
+        id: this.id,
+        role: this.messageRole,
+        content: this.content,
+      },
+    ];
+  }
+
+  async put(
+    _messages: MemoryMessage<TAdditionalMessageOptions>[],
+  ): Promise<void> {
+    // No-op: static content doesn't change
+  }
+}
@@ -1,13 +1,14 @@
-import { Settings } from "../global";
-import type { ChatMessage } from "../llms";
-import { type BaseChatStore, SimpleChatStore } from "../storage/chat-store";
-import { extractText } from "../utils";
+import { Settings } from "../../global";
+import type { ChatMessage } from "../../llms";
+import { type BaseChatStore, SimpleChatStore } from "../../storage/chat-store";
+import { extractText } from "../../utils";

 export const DEFAULT_TOKEN_LIMIT_RATIO = 0.75;
 export const DEFAULT_CHAT_STORE_KEY = "chat_history";

 /**
 * A ChatMemory is used to keep the state of back and forth chat messages
+ * @deprecated Use Memory instead.
 */
 export abstract class BaseMemory<
  AdditionalMessageOptions extends object = object,
@@ -55,6 +56,9 @@ export abstract class BaseMemory<
  }
 }

+/**
+ * @deprecated Use Memory with snapshot feature with your own storage instead.
+ */
 export abstract class BaseChatStoreMemory<
  AdditionalMessageOptions extends object = object,
 > extends BaseMemory<AdditionalMessageOptions> {
@@ -1,6 +1,6 @@
-import { Settings } from "../global";
-import type { ChatMessage, LLM } from "../llms";
-import { type BaseChatStore } from "../storage/chat-store";
+import { Settings } from "../../global";
+import type { ChatMessage, LLM } from "../../llms";
+import { type BaseChatStore } from "../../storage/chat-store";
 import { BaseChatStoreMemory, DEFAULT_TOKEN_LIMIT_RATIO } from "./base";

 type ChatMemoryBufferOptions<AdditionalMessageOptions extends object = object> =
@@ -12,6 +12,9 @@ type ChatMemoryBufferOptions<AdditionalMessageOptions extends object = object> =
    llm?: LLM<object, AdditionalMessageOptions> | undefined;
  };

+/**
+ * @deprecated Use Memory instead.
+ */
 export class ChatMemoryBuffer<
  AdditionalMessageOptions extends object = object,
 > extends BaseChatStoreMemory<AdditionalMessageOptions> {
@@ -1,10 +1,13 @@
 import { type Tokenizer, tokenizers } from "@llamaindex/env/tokenizers";
-import { Settings } from "../global";
-import type { ChatMessage, LLM, MessageType } from "../llms";
-import { defaultSummaryPrompt, type SummaryPrompt } from "../prompts";
-import { extractText, messagesToHistory } from "../utils";
+import { Settings } from "../../global";
+import type { ChatMessage, LLM, MessageType } from "../../llms";
+import { defaultSummaryPrompt, type SummaryPrompt } from "../../prompts";
+import { extractText, messagesToHistory } from "../../utils";
 import { BaseMemory } from "./base";

+/**
+ * @deprecated Use Memory instead.
+ */
 export class ChatSummaryMemoryBuffer extends BaseMemory {
  /**
   * Tokenizer function that converts text to tokens,
@@ -0,0 +1,136 @@
+import type { ChatMessage } from "../llms";
+import { ChatMessageAdapter } from "./adapter/chat";
+import {
+  FactExtractionMemoryBlock,
+  type FactExtractionMemoryBlockOptions,
+} from "./block/fact";
+import {
+  StaticMemoryBlock,
+  type StaticMemoryBlockOptions,
+} from "./block/static";
+import { DEFAULT_TOKEN_LIMIT, Memory, type MemoryOptions } from "./memory";
+import type { MemoryMessage } from "./types";
+
+/**
+ * Create a Memory instance with default options
+ * @returns A new Memory instance with default configuration
+ */
+export function createMemory<TMessageOptions extends object = object>(): Memory<
+  Record<string, never>,
+  TMessageOptions
+>;
+
+/**
+ * Create a Memory instance with options only
+ * @param options - Memory configuration options
+ * @returns A new Memory instance
+ */
+export function createMemory<TMessageOptions extends object = object>(
+  options: MemoryOptions<TMessageOptions>,
+): Memory<Record<string, never>, TMessageOptions>;
+
+/**
+ * Create a Memory instance with ChatMessage array (IDs will be generated)
+ * @param messages - Initial ChatMessage array for the memory
+ * @param options - Memory configuration options
+ * @returns A new Memory instance
+ */
+export function createMemory<TMessageOptions extends object = object>(
+  messages: ChatMessage<TMessageOptions>[],
+  options?: MemoryOptions<TMessageOptions>,
+): Memory<Record<string, never>, TMessageOptions>;
+
+/**
+ * Create a Memory instance with MemoryMessage array and options
+ * @param messages - Initial MemoryMessage array for the memory
+ * @param options - Memory configuration options
+ * @returns A new Memory instance
+ */
+export function createMemory<TMessageOptions extends object = object>(
+  messages: MemoryMessage<TMessageOptions>[],
+  options: MemoryOptions<TMessageOptions>,
+): Memory<Record<string, never>, TMessageOptions>;
+
+/**
+ * Create a Memory instance
+ * @param messagesOrOptions - Either initial messages or options
+ * @param options - Memory configuration options (when first param is messages)
+ * @returns A new Memory instance
+ */
+export function createMemory<TMessageOptions extends object = object>(
+  messagesOrOptions:
+    | ChatMessage<TMessageOptions>[]
+    | MemoryMessage<TMessageOptions>[]
+    | MemoryOptions<TMessageOptions> = [],
+  options: MemoryOptions<TMessageOptions> = {},
+): Memory<Record<string, never>, TMessageOptions> {
+  let messages: MemoryMessage<TMessageOptions>[] = [];
+
+  if (Array.isArray(messagesOrOptions)) {
+    const firstMessage = messagesOrOptions[0];
+    if (firstMessage) {
+      if ("id" in firstMessage) {
+        messages = messagesOrOptions as MemoryMessage<TMessageOptions>[];
+      } else {
+        const adapter = new ChatMessageAdapter<TMessageOptions>();
+        messages = messagesOrOptions.map((chatMessage) =>
+          adapter.toMemory(chatMessage),
+        );
+      }
+    }
+  }
+  return new Memory<Record<string, never>, TMessageOptions>(messages, options);
+}
+
+/**
+ * create a StaticMemoryBlock
+ * @param options - Configuration options for the static memory block
+ * @returns A new StaticMemoryBlock instance
+ */
+export function staticBlock<TMessageOptions extends object = object>(
+  options: StaticMemoryBlockOptions,
+): StaticMemoryBlock<TMessageOptions> {
+  return new StaticMemoryBlock<TMessageOptions>(options);
+}
+
+/**
+ * create a FactExtractionMemoryBlock
+ * @param options - Configuration options for the fact extraction memory block
+ * @returns A new FactExtractionMemoryBlock instance
+ */
+export function factExtractionBlock<TMessageOptions extends object = object>(
+  options: FactExtractionMemoryBlockOptions,
+): FactExtractionMemoryBlock<TMessageOptions> {
+  return new FactExtractionMemoryBlock<TMessageOptions>(options);
+}
+
+/**
+ * Creates a new Memory instance from a snapshot
+ * @param snapshot The snapshot to load from
+ * @param options Optional MemoryOptions to apply when loading (including memory blocks)
+ * @returns A new Memory instance with the snapshot data and provided options
+ */
+export function loadMemory<TMessageOptions extends object = object>(
+  snapshot: string,
+  options?: MemoryOptions<TMessageOptions>,
+): Memory<Record<string, never>, TMessageOptions> {
+  const { messages, tokenLimit, memoryCursor } = JSON.parse(snapshot);
+
+  // Merge snapshot data with provided options
+  const mergedOptions: MemoryOptions<TMessageOptions> = {
+    tokenLimit: options?.tokenLimit ?? tokenLimit ?? DEFAULT_TOKEN_LIMIT,
+    ...(options?.shortTermTokenLimitRatio && {
+      shortTermTokenLimitRatio: options.shortTermTokenLimitRatio,
+    }),
+    ...(options?.customAdapters && {
+      customAdapters: options.customAdapters,
+    }),
+    memoryBlocks: options?.memoryBlocks ?? [],
+    memoryCursor: memoryCursor ?? 0,
+  };
+
+  return new Memory<Record<string, never>, TMessageOptions>(
+    messages,
+    mergedOptions,
+  );
+}
@@ -1,3 +1,9 @@
-export { BaseMemory } from "./base";
-export { ChatMemoryBuffer } from "./chat-memory-buffer";
-export { ChatSummaryMemoryBuffer } from "./summary-memory";
+export { BaseMemory } from "./deprecated/base";
+export { ChatMemoryBuffer } from "./deprecated/chat-memory-buffer";
+export { ChatSummaryMemoryBuffer } from "./deprecated/summary-memory";
+
+export * from "./adapter";
+export * from "./block";
+export * from "./factories";
+export { Memory } from "./memory";
+export * from "./types";
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
github-actions[bot]	d68c2a4be8	Release 0.11.13 (#2060 )	2025-07-07 11:24:21 +07:00
Alex Yang	47a7555c07	chore: bump sdk version (#2062 )	2025-07-03 12:05:16 -07:00
Marcus Schiesser	363bfa778e	chore: re-add lib folder from docs and rename it to libs (so pnpm clean doesn't delete it)	2025-07-03 11:03:05 +07:00
Jan Z	229cdeb0ff	feat: add agent update to groq models (#2054 ) Co-authored-by: Marcus Schiesser <marcus.schiesser@googlemail.com> Co-authored-by: Marcus Schiesser <mail@marcusschiesser.de>	2025-07-01 22:53:47 -07:00
github-actions[bot]	7a2485cca2	Release 0.11.12 (#2050 ) Co-authored-by: marcusschiesser <17126+marcusschiesser@users.noreply.github.com>	2025-07-02 11:41:55 +07:00
Marcus Schiesser	1329186a23	docs: clarify how to run docs	2025-07-02 11:33:48 +07:00
dependabot[bot]	5d6e7384f5	chore(deps-dev): bump @modelcontextprotocol/server-filesystem from 2025.3.28 to 2025.7.1 (#2055 )	2025-07-02 11:26:18 +07:00
allen	f2dfd305fb	implement bm25 retriever (#2045 ) Co-authored-by: Marcus Schiesser <mail@marcusschiesser.de>	2025-07-02 11:22:47 +07:00
Huu Le	3cd8a573df	feat: update interpreter to always upload all files in the configured directory (#2057 )	2025-07-02 10:57:04 +07:00
Laurie Voss	09c6077f6e	Import path for llamaparsereader (#2056 )	2025-07-01 16:51:25 -07:00
Logan	14cc65b4e3	add google analytics (#2053 ) Co-authored-by: Alex Yang <himself65@outlook.com>	2025-07-01 11:18:14 -07:00
Marcus Schiesser	c544d8f67c	docs: review and update memory doc	2025-07-01 15:10:43 +07:00
Huu Le	d578889e21	feat: new memory api (#2028 ) Co-authored-by: Marcus Schiesser <mail@marcusschiesser.de>	2025-07-01 09:30:49 +07:00
Marcus Schiesser	9f745d1941	chore: revert to wrong opus change	2025-07-01 09:07:46 +07:00
Alex Yang	f292e94dcd	fix: change default claude model (#2052 )	2025-06-30 15:19:40 -07:00
Marcus Schiesser	0fcc92f632	fix: sentence splitter must not trim whitespaces (#2046 )	2025-06-30 17:32:04 +07:00
Marcus Schiesser	515a8b9111	fix: error logging for fromPersistPath (#2049 )	2025-06-30 13:41:13 +07:00
github-actions[bot]	7e8efc6284	Release @llamaindex/tools@0.1.2 (#2048 )	2025-06-30 11:40:54 +07:00
Wassim Chegham	0fcf65126d	chore: export type MCPClientOptions (#2047 ) Co-authored-by: Marcus Schiesser <marcus.schiesser@googlemail.com>	2025-06-28 10:55:07 +07:00