Ce contenu n’est pas encore disponible dans votre langue.
A workflow is a chain of nodes: a trigger starts it, and each node does one step. Type what you want to do, such as “click”, “pdf” or “summarise”, or narrow the list by family, by where the node runs and by what it needs before it works. Each card opens the node’s page: its settings, what it sends to the next node, and fixes for common problems.
Famille
Fonctionne dans
Nécessite
Seulement
Affichage de 189 nœuds
Déclencheurs 10
- Browser Startup Run the flow when the browser starts. Aucune configurationExtension uniquement
- Chat Trigger Run the flow on each chat message from the user. Aucune configuration
- Context Menu Run the flow from the browser context menu. Aucune configurationExtension uniquement
- Element Change Run the flow when any of one or more CSS selectors on a matching page changes. Aucune configurationExtension uniquement
- Error Trigger Run the flow when a workflow fails. Aucune configurationExtension uniquement
- Hotkey Run the flow from a keyboard shortcut. Aucune configurationExtension uniquement
- Navigation Run the flow when a browser tab is created, updated, or removed. Aucune configurationExtension uniquement
- Page Load Run the flow when a matching browser tab finishes loading. Aucune configurationExtension uniquement
- Schedule Run the flow at a specific time. Aucune configuration
- When Started Run the flow on clicking the start button. Good for getting started quickly. Aucune configuration
Dans la page 27
- Browser Notification Show a system notification through the OS notification center, visible even when no tab is focused. Unlike Show Notification, which needs an open tab, this reaches the user regardless of what they are looking at, and can offer up to two action buttons. Aucune configurationExtension uniquement
- Clipboard Write Write text to the system clipboard on the active tab. This node never reads the clipboard; reading clipboard content is a separate, permission-gated node. Aucune configurationExtension uniquement
- Delete UI Elements Delete selected UI elements from the opened web page Aucune configurationExtension uniquement
- Display Markdown Display markdown content in the UI Aucune configurationExtension uniquement
- Display Object Display an object in the UI Aucune configurationExtension uniquement
- Form Fill Fill a form on the opened web page Aucune configurationExtension uniquement
- Get Accessibility Tree Get visible interactive and semantic elements from the opened page Aucune configurationExtension uniquement
- Get All HTML Get all the html from the page of the actual openend tab Aucune configurationExtension uniquement
- Get All Images Get all the images from the page of the actual openend tab Aucune configurationExtension uniquement
- Get All Link Get all the links from the page of the actual openend tab Aucune configurationExtension uniquement
- Get All Text Get the selected text from the page of the actual openend tab Aucune configurationExtension uniquement
- Get Element Data Get details for a selected element in the opened page Aucune configurationExtension uniquement
- Get Forms Schema Get schemas for one form or all forms from the opened tab Aucune configurationExtension uniquement
- Get HTML of Selected Text Get the selected text from the page of the actual openend tab Aucune configurationExtension uniquement
- Get Page Assets Get asset and downloadable link URLs from the opened page Aucune configurationExtension uniquement
- Get Page Metadata Get metadata from the opened page Aucune configurationExtension uniquement
- Get Selected Text Get the selected text from the page of the actual openend tab Aucune configurationExtension uniquement
- Get Structured Data Extract structured data from a table or list in the opened tab Aucune configurationExtension uniquement
- Highlight Elements Highlight selected elements in the opened page Aucune configurationExtension uniquement
- Screenshot Page Capture a screenshot of the opened page or selected element Aucune configurationExtension uniquement
- Set Element Attribute Style Set attributes and styles on selected elements in the opened page Aucune configurationExtension uniquement
- Show Notification Show a native browser notification You can use this node when a background or browser-triggered workflow needs to notify the user.. Aucune configurationExtension uniquement
- Submit Form Submit a selected form in the opened page Aucune configurationExtension uniquement
- Tab Control Open, close, focus, reload, navigate, or group browser tabs. Operates on the active tab when no tab id is given. Aucune configurationExtension uniquement
- UI Automation Automate UI interactions Aucune configurationExtension uniquement
- Wait For Element Wait for an element condition in the opened page Aucune configurationExtension uniquement
- Web Search Run a web search and return the results as JSON Aucune configurationExtension uniquement
Flux 18
- Compare Datasets Sort two collections into new, gone, unchanged and changed Aucune configuration
- Filter Accept items matching the conditions Aucune configuration
- Funnel Continues whenever any one of its incoming branches completes. Use it to converge multiple triggers (or parallel branches) into a single shared downstream chain. Aucune configuration
- Get Variable Emit a stored value as the item Aucune configuration
- IF Route items to different branches based on conditions Aucune configuration
- In-Page Form Pause the run and ask the user to fill a form; continues with the values they enter. Aucune configuration
- Loop Execute a set of actions repeatedly based on conditions Aucune configuration
- Merge Merge multiple streams into one Aucune configuration
- No Operation Pass data through unchanged Aucune configuration
- Poll Until Call an endpoint on an interval until a condition is met, or give up Aucune configuration
- Record Audio Pause the run and record audio from the microphone; continue with the recording. Aucune configuration
- Set Variable Remember a value for the rest of this run Aucune configuration
- Split in Batches Loop over a list a batch at a time, running the body once per batch Aucune configuration
- Stop and Error Stop the workflow and throw an error Aucune configuration
- Switch Route items to any number of named branches Aucune configuration
- Try / Catch Guard the work after it, and route a failure instead of ending the run Aucune configuration
- Wait Pause the workflow for a specified amount of time Aucune configuration
- Wait For Approval Pause the run until a person approves or rejects it Aucune configuration
Données 32
- Add To A Date Add a number of days to a date Aucune configuration
- Aggregate Combine many items into a single item Aucune configuration
- Compression Compress or decompress binary data with gzip Aucune configuration
- Convert to File Turn items into a file (CSV, JSON, text, base64 or binary) Aucune configuration
- Crypto Hash, HMAC and generate random values with WebCrypto Aucune configuration
- CSV Convert between CSV text and items Aucune configuration
- Download as file Download the input data as a file Aucune configuration
- Edit Fields Modify, add or remove item fields Aucune configuration
- Edit Image Resize, crop, rotate, convert, compress or annotate an image Aucune configuration
- Extract From File Turn a file (CSV, JSON, XML, text or PDF) into items, format auto-detected Aucune configuration
- Extract Part Of A Date Extract a part of a date Aucune configuration
- Format Date Format a date Aucune configuration
- Get Current Date Get the current date Aucune configuration
- Get Time Between Dates Get the time between two dates Aucune configuration
- HTML Extract Pull values out of HTML with CSS selectors Aucune configuration
- JSON Parse, build, pick from and repair JSON Aucune configuration
- Limit Limit the number of items to a maximum count Aucune configuration
- Local File Read Opens a file picker at runtime and reads the chosen file(s). The file is selected by the user through the operating system picker, so the workflow can never name a file itself. Aucune configuration
- Markdown Convert HTML to Markdown, or Markdown to HTML Aucune configuration
- Math Calculate with numbers and format them for people Aucune configuration
- PDF Parser Extract text from a PDF document and output it as a string Aucune configuration
- Pick Field Pick a specific field from the input data Aucune configuration
- Regex Match, extract and replace with a regular expression Aucune configuration
- Remove Duplicates Remove repeated items from a collection Aucune configuration
- Rename Keys Rename and move keys on an item Aucune configuration
- Round a Date Round a date up or down to a chosen unit Aucune configuration
- Save as File Save text content as a downloadable file Aucune configuration
- Sort Sort items by one or more fields, or shuffle them randomly Aucune configuration
- Split Out Turn a list inside an item into separate items Aucune configuration
- Subtract From A Date Subtract a number of days from a date Aucune configuration
- Text / Formatter Transform text: case, trim, pad, replace, split, slugify, extract, count, encode and decode Aucune configuration
- XML Convert between XML and JSON Aucune configuration
Cœur 11
- Code Run custom python code Aucune configuration
- Data Store Remember values between workflow runs Aucune configuration
- Debug Generate sample data, delay, or throw on demand Aucune configuration
- Digest Accumulate items across runs and release them together Aucune configuration
- Execution Data Read this run’s id, trigger and timing, or attach searchable metadata to it Aucune configuration
- Get All Text from Link Get all the text content from a given URL Aucune configuration
- Get HTML from Link Get the HTML content from a given URL Aucune configuration
- Get Images from Link Get the Images content of a page from a given URL Aucune configuration
- Get Links from Link Get the Links content of a page from a given URL Aucune configuration
- Http Request Makes a HTTP request and returns the response data Aucune configuration
- RSS Feed Read Reads an RSS feed from a URL and returns the parsed data Aucune configuration
IA 65
- Analyze Text On-device sentiment / toxicity / moderation of text Sur l'appareil
- Answer Question Extract an answer from a passage, on-device Sur l'appareil
- Auto-fixing Output Parser Wraps another output parser and retries with a model when the output fails to parse Aucune configuration
- Basic LLM Chain A simple chain to prompt a LLM and return the response Aucune configuration
- Browser Tool Lets the model click/fill/type/select/submit on the active page or navigate it. Each call requires user approval. Aucune configurationExtension uniquementOutil d'agent
- Calculator Evaluate arithmetic expressions. Aucune configurationOutil d'agent
- Caption Image Describe an image in words, on-device Sur l'appareil
- Character Text Splitter Breaks down text into smaller parts. Aucune configuration
- Chat Anthropic Use Anthropic Claude chat models through the Anthropic API. Nécessite une connexion
- Chat Azure OpenAI Use OpenAI models hosted on Azure OpenAI Service. Nécessite une connexion
- Chat DeepSeek Use DeepSeek chat and reasoning models. Nécessite une connexion
- Chat Google Use Google Gemini chat models through the Gemini API. Nécessite une connexion
- Chat Groq Run open models on Groq for fast, low-latency inference. Nécessite une connexion
- Chat Memory Lets an agent remember the conversation, across runs or within one run NouveauAucune configuration
- Chat Mistral Run Mistral open and frontier models for text generation. Nécessite une connexion
- Chat OpenAI Use OpenAI chat models (GPT) through the OpenAI API. Nécessite une connexion
- Chat OpenRouter Route to hundreds of models across providers through OpenRouter. Nécessite une connexion
- Chat xAI (Grok) Use xAI Grok chat models. Nécessite une connexion
- Chrome AI Leverage Chrome AI to use LLM directly on your device Sur l'appareil
- Classify Image Label an image on-device with any installed image-classification model Sur l'appareil
- Classify Text Sort text into your own labels on-device — no training needed Sur l'appareil
- Cohere Embeddings Leverage Cohere to create embeddings. Nécessite une connexion
- Current Page Document Loader Load the current tab as text to feed into the Indexer Aucune configurationExtension uniquement
- Detect Objects Find objects + bounding boxes in an image, on-device Sur l'appareil
- File Document Loader Load a file (PDF, HTML, text) as text to feed into the Indexer Aucune configuration
- Generate Text Summarize, translate or transform text on-device Sur l'appareil
- Google Embeddings Leverage Google (Gemini) to create embeddings. Nécessite une connexion
- Guardrails Validate text against policies using a model, and route to pass or fail Aucune configuration
- HTTP Request Tool Lets the model send HTTP requests. Each call requires user approval. Aucune configurationOutil d'agent
- Hugging Face Embeddings Leverage Hugging Face Inference to create embeddings. Nécessite une connexion
- Indexer Splitting text into chunks and embedding them into a vector store Aucune configuration
- Information Extractor Extract structured information from a text using a LLM Aucune configuration
- Item List Output Parser Parse a comma-separated list of items from the model output Aucune configuration
- Local Classifier Classify text on-device: zero-shot, sentiment or toxicity Sur l'appareil
- Local Embeddings Turn text into vectors on this device — free and private Sur l'appareil
- Local Knowledge Gives the connected agent a knowledge base to search Aucune configuration
- Local Q&A Model Answer questions on-device by quoting your Knowledge Sur l'appareil
- MCP Client Tool Connects to a remote MCP server and lets the model call its tools. Each call requires user approval. Aucune configurationOutil d'agent
- Ollama Use models served by Ollama on your own computer. Sur l'appareil
- Ollama Embeddings Create embeddings with models served by Ollama on your own computer. Sur l'appareil
- OpenAI Embeddings Leverage OpenAI to create embeddings. Nécessite une connexion
- Q&A Agent Answer a question from the connected knowledge base, in one step. Aucune configuration
- RAG Agent Answer from your knowledge base: finds the relevant passages and replies with sources, remembering the conversation. Aucune configuration
- Recursive Character Text Splitter Breaks down text into smaller parts. Aucune configuration
- Run agent Give a task to one of your agents and get its answer Aucune configuration
- Run Local Model Run any installed on-device model for a chosen task Sur l'appareil
- Sentiment Analysis Analyze the sentiment of a text using a LLM, or on-device with a Local Classifier Aucune configuration
- Structured Output Parser Return data in a defined JSON format Aucune configuration
- Summarization Summarize a text using a LLM Aucune configuration
- Summarize Condense long text on-device Sur l'appareil
- Text Classifier Classify text into one of a set of given categories using a LLM, or on-device with a Local Classifier Aucune configuration
- Think Give the agent a scratchpad to reason through a problem step by step. Aucune configurationOutil d'agent
- Tools Agent Generate an action plan and execute it. Can use external tools. Aucune configuration
- Transcribe Audio Speech-to-text on-device (Whisper) — transcript + timestamped segments Sur l'appareil
- Transformers Chat Run a small chat model on-device with transformers.js Sur l'appareil
- Translate Translate text between 200 languages on-device Sur l'appareil
- URL Document Loader Load a web page as text to feed into the Indexer Aucune configuration
- Voyage Embeddings Leverage Voyage AI to create embeddings. Nécessite une connexion
- Web LLM Use LLM locally directly on your device Sur l'appareil
- Web Search Search the web and return results. Aucune configurationExtension uniquementOutil d'agent
- Wikipedia Query Query Wikipedia for information on a given topic. Aucune configurationOutil d'agent
- Workflow as Tool Expose another workflow as a callable tool. Aucune configurationOutil d'agent
- Local Memory Replaced by Chat Memory. Existing workflows keep running it unchanged. ObsolèteAucune configuration
- Persistent Chat Memory Replaced by Chat Memory (key storage, shared scope). Existing workflows keep running it unchanged. ObsolèteAucune configuration
- Window Buffer Memory Replaced by Chat Memory ("This run only" storage). Existing workflows keep running it. ObsolèteAucune configuration
Intégrations 24
Chat
- Discord Integrate with Discord: post messages and rich embeds via a webhook, or read channels and messages with a bot token. Nécessite une connexion6 opérationsExtension uniquement
- Slack Integrate with Slack to perform operations on channels, messages, users, and files. Nécessite une connexion11 opérationsExtension uniquement
- Telegram Send messages, photos and documents to a Telegram chat via the Bot API. Nécessite une connexion3 opérations
Notes & docs
- Google Drive Integrate with Google Drive to manage files and folders. Nécessite une connexion10 opérations
- Notion Integrate with Notion to perform operations on databases, pages, and blocks. Nécessite une connexion11 opérationsExtension uniquement
- Obsidian Append, create, read, and search notes in a local Obsidian vault via the Local REST API plugin. Nécessite une connexion4 opérationsExtension uniquement
Tasks
- Linear Create and update Linear issues, add comments, and list teams and projects — with the current page as context. Nécessite une connexion7 opérationsExtension uniquement
- Todoist Create and manage Todoist tasks from the current page, with natural-language due dates. Nécessite une connexion6 opérationsExtension uniquement
- Trello Create, update, and move Trello cards from the current page, and attach the page URL. Nécessite une connexion6 opérationsExtension uniquement
Sheets & databases
- AirTable Integrate with AirTable to perform operations on tables. Nécessite une connexion7 opérations
- Baserow / NocoDB Create, read, update, and delete rows in a Baserow or NocoDB table — hosted or self-hosted. Nécessite une connexion5 opérationsExtension uniquement
- Google Sheets Integrate with Google Sheets to perform operations on spreadsheets and data. Nécessite une connexion16 opérations
- Supabase Select, insert, update, upsert, and delete rows in a Supabase project via PostgREST, plus RPC calls. Nécessite une connexion6 opérationsExtension uniquement
CRM
- Attio Create or update people and companies, and log notes from the current page in Attio CRM. Nécessite une connexion5 opérationsExtension uniquement
- HubSpot Create or update contacts and companies, log notes from the current page, and list deals in HubSpot CRM. Nécessite une connexion7 opérationsExtension uniquement
- Pipedrive Create or update persons and organizations, and log notes from the current page in Pipedrive CRM. Nécessite une connexion5 opérationsExtension uniquement
Read-later
- Raindrop Save the current page to a Raindrop collection with title, excerpt, tags and cover — with duplicate-skip. Nécessite une connexion5 opérationsExtension uniquement
- Readwise / Reader Save highlights to Readwise (batch) and full articles to Reader, carrying source URL, title and author. Nécessite une connexion2 opérationsExtension uniquement
Dev
- Custom API Call any REST API from its OpenAPI (Swagger) spec — pick an operation and its fields are generated for you. Aucune configurationExtension uniquement
- GitHub Integrate with GitHub to work with issues, pull requests, repositories, and releases. Nécessite une connexion16 opérations
Other
- Gmail Integrate with Gmail to perform operations on messages and drafts. Nécessite une connexion9 opérations
- Google Calendar Integrate with Google Calendar to create, list and update events. Nécessite une connexion7 opérations
- Microsoft 365 Integrate with Microsoft 365 via Graph: send and read Outlook mail, manage OneDrive files, create calendar events and To Do tasks. Nécessite une connexion11 opérationsExtension uniquement
- Push Notification Send a push notification to your phone via ntfy or Pushover, with priority and a click-through URL. Aucune configurationExtension uniquement
Lambda 2
Start with a goal
Section titled “Start with a goal”| Goal | Start with |
|---|---|
| Start by hand | When Started |
| Start on a schedule | Schedule |
| Start from the browser | Context Menu, Hotkey or Page Load |
| Start on each chat message | Chat Trigger |
| Read the page you’re on | Get Selected Text, Get All Text or Get Structured Data |
| Read a web address | Get HTML from Link or Get All Text from Link |
| Call an API | HTTP Request |
| Reshape data | Edit Fields or Pick Field |
| Branch, filter, merge, loop or wait | IF, Filter, Merge, Loop, Wait |
| Process many items without hitting rate limits | Split in Batches (see Rate limits) |
| Pause for a person’s approval | Wait For Approval |
| Ask an AI model | Basic LLM Chain with a chat model such as Chat OpenAI, or on-device with Summarize |
| Let an AI agent pick tools | Tools Agent |
| Send results to an app | Slack, Notion, Google Sheets and the other integrations above |
How to read a node page
Section titled “How to read a node page”Every node page follows the same order, so you know where to look:
- Needs a connection: when the node signs in to a service, a box at the top links to the connection setup.
- Try it: a real workflow that uses the node, which you can open in AWFlow.
- What it does and When to use it.
- Inputs and settings: every field as the settings panel shows it. For apps with many operations, pick one to see only its settings.
- Outputs: the output ports and the fields the next node receives, with the shape of one item.
- Dependencies and credentials: the connection the node uses and the slots where other nodes plug in, such as a chat model for an agent.
- Example workflow, Common issues (named by what you see), Recipes with this node and Related nodes.
Nodes that act on web pages or the browser itself run only in the AWFlow browser extension. Their pages say so at the top.