Browse 160,118 AI agents

By Zayco3D

Quick Overview This workflow uses Google Sheets and a Telegram bot to send scheduled follow-ups for pending quotes, record each follow-up stage back in the sheet, send an optional daily digest to the owner, and classify incoming Telegram replies to notify you or mark quotes as lost. How it works Runs every day at 09:00 based on a schedule trigger. Reads all rows from the “Quotes” tab in Google Sheets and evaluates each pending quote’s age and last follow-up stage to decide whether a stage 1, 2, or 3 follow-up is due (and avoids sending twice on the same day). Sends the appropriate personalized follow-up message to each customer via Telegram and updates the Google Sheet with the follow-up stage and date. Builds a daily digest summarizing how many follow-ups were sent per stage, plus any skipped rows or quotes pending for more than 30 days. If the digest is enabled and an owner chat ID is set, sends the digest to the owner via Telegram. Separately, triggers on incoming Telegram messages, classifies each reply as positive/negative/neutral, notifies the owner about positive or neutral replies, and marks the sender’s pending quote as lost in Google Sheets (then notifies the owner) for negative replies. Setup Connect your Google Sheets OAuth credentials and set the spreadsheet Document ID for all Google Sheets nodes that read/update the “Quotes” tab. Create a Telegram bot with BotFather and add the bot token to both Telegram credentials (Telegram and Telegram Trigger). Create a “Quotes” sheet with columns: Name, ChatID, Item, Amount, QuoteDate (YYYY-MM-DD), Status, LastFollowUp, and LastFollowUpDate, and ensure pending quotes use Status = "pending". In “Set Follow-Up Config”, set your follow-up day thresholds, message templates, whether the digest is enabled, and your owner_chat_id for the daily digest. In “Set Owner Config”, set the owner_chat_id that should receive reply alerts (you can get your chat ID via @userinfobot). An n8n automation workflow template by Zayco3D.

By Devdesk App

Quick overview This workflow runs on a manual test trigger to sanitize an incoming payment webhook payload using DevDeskApp, generating a prettified JSON body, a SHA-256 checksum, and a UUID v4 trace ID, then consolidating everything into a database-ready audit record. How it works Click the orange 'Test workflow' button at the bottom of the canvas to execute the complete zero-trust webhook ingestion pipeline. The sample incoming raw payment payload passes through the DevDeskApp JSON Prettifier, Cryptographic SHA-256 Hasher, and Distributed Trace UUID Generator. Inspect the final output node to review the sanitized database record and the authenticated DevOps Slack alert without needing any external credentials. Creates a sample payment webhook event containing headers and a raw, stringified JSON body. Uses DevDeskApp to validate and prettify the raw JSON payload into a formatted JSON string. Uses DevDeskApp to generate a SHA-256 hash of the formatted payload for integrity auditing. Uses DevDeskApp to generate an RFC UUID v4 trace ID for correlating the event across systems. Consolidates the parsed payload, checksum, trace ID, and key metadata into a single “audit log” object suitable for database storage. Setup Install and enable the DevDeskApp n8n community node (n8n-nodes-devdeskapp) used for JSON formatting, hashing, and UUID generation. Replace the sample payload generator with a real n8n Webhook trigger and map the incoming raw body and headers to the DevDeskApp JSON prettifier input. Update the final consolidated audit object (for example, table name and fields) to match your database schema and add a database or alerting node if you want to persist or notify. An n8n automation workflow template by Devdesk App.

By Glasser.AI

Quick overview Find existing pages worth refreshing with Google Search Console, research competing pages through Glasser, and generate source-backed Claude editing tasks in Google Sheets. Record edits and compare equal before-and-after windows on later runs. How it works Runs on a manual trigger or on a weekly schedule every Monday. Pulls two adjacent 28-day performance windows from the Google Search Console Search Analytics API for a configured property and country. Reads the existing task ledger from Google Sheets and selects up to a set number of non-brand page/query candidates in a target position and impression range, while also identifying previously modified tasks that are due for review. For each new candidate, uses Glasser to fetch SERP results (Serper provider), selects competitor URLs ahead of your page, and scrapes your page plus competitor pages (Firecrawl provider) for readable content. If enough own and competitor evidence is available, sends the page evidence to Anthropic Claude to produce a JSON content brief with 1–5 evidence-backed edit actions, otherwise flags the task as needing manual review. Appends or updates the task row in Google Sheets, preserving existing tasks by task_id. For due reviews, queries Google Search Console for equal 28-day windows before and after the recorded edit date and updates the same Google Sheets row with the comparison metrics and review status. Setup Create a Google Sheet with a Tasks tab. Copy the tab-separated header line from the Task sheet setup sticky note into A1, then set spreadsheetId in Settings. Keep every header unchanged. Add Google OAuth2 credentials with the Search Console API enabled and scope https://www.googleapis.com/auth/webmasters.readonly, and ensure the Google account has access to the configured Search Console property. Add Google Sheets OAuth credentials and connect them to the Google Sheets nodes that read and append/update the task ledger. Install and configure the n8n-nodes-glasser community node and provide Glasser credentials with access to SERP (Serper) and scraping (Firecrawl) providers. Add Anthropic (Claude) credentials (or n8n AI Gateway access) and select an available model ID in Settings. Review and adjust Settings values such as domain/property, countries, brand terms, position/impression thresholds, maxPages/maxReviews, and rivalsPerPage before enabling the scheduled trigger. Requirements An n8n instance that permits n8n-nodes-glasser; a Glasser API key with balance; a Google Search Console property and read-only OAuth access; Google Sheets edit access; and Anthropic credentials or supported n8n Cloud Gateway credits. Glasser and model calls incur usage charges. Customization Configure your domain, property, brand terms, country codes, position band, minimum impressions, page and review limits, competitor count, product context and an available Claude model in Settings. Start with maxPages=1. Additional info Website edits remain manual. After an edit, set status=modified and modified_at=YYYY-MM-DD using the Pacific date; reviews wait for a complete 28-day window plus a three-day buffer. Comparisons are observational, and missing data remains unknown. Live validation used a real GSC CSV export to test Glasser, Claude and Sheets. GSC OAuth API access and the post-edit review branch still require live validation. An n8n automation workflow template by Glasser.AI.

By Ali Abdullah

Quick overview This workflow automates return merchandise authorizations (RMAs) for Shopify and WooCommerce by ingesting return requests, fetching order details, classifying return reasons with Google Gemini, checking prior return history in Google Sheets, then either auto-approving with customer/warehouse emails or routing to Slack for human approval. How it works Receives a return request via a webhook, runs hourly to pull Shopify orders tagged return-requested, or starts from a manual test trigger. Normalizes the intake payload and fetches the matching order from Shopify or WooCommerce to calculate order value, fulfillment status, and order age. Sends the customer’s reason text and item details to Google Gemini to classify the return reason into a predefined category with a confidence score. Reads the RMALog tab in Google Sheets to count the customer’s recent returns and flags the request for potential fraud review. Generates a unique RMA ID, assigns an inventory disposition (restock/quarantine/damaged) per item, and decides whether the return can be auto-approved based on value, category, fulfillment, time window, and fraud signals. If auto-approved (or later approved via the decision webhook), creates the return/refund in Shopify or WooCommerce, emails a shipping-label request to the warehouse via Gmail, emails the approval to the customer via Gmail, and appends the RMA record to Google Sheets. If human review is required, posts the RMA summary to Slack and logs a pending row in Google Sheets, then a separate approval webhook accepts an approved/rejected decision to either proceed with platform return creation or email the customer a rejection. Setup Add credentials for Shopify (Admin API access token) and/or WooCommerce (REST API consumer key/secret) and fill in your store URL/domain values in the workflow configuration. Provide a Google Gemini API key in the workflow configuration to enable return-reason classification. Connect Google Sheets credentials, set the RMA spreadsheet ID, and ensure the document has an RMALog sheet with columns matching the workflow’s appended fields (including date and customer_email). Connect Gmail credentials and set the warehouse email address and customer email sending permissions for your mailbox. Connect Slack credentials and set the target channel ID for ops review notifications. Copy the two webhook URLs (/return-request and /return-decision) and configure your return form/helpdesk and approval source to POST the expected fields (including order_id, platform, and rma_id/decision). An n8n automation workflow template by Ali Abdullah.

By Ali Abdullah

Quick overview This workflow runs monthly to collect SOC 2 evidence from GitHub, AWS IAM, and Google Workspace, logs checks to a Google Sheets evidence ledger, generates a summary in Google Docs, exports it to PDF in Google Drive, emails it via Gmail, and posts a Slack alert if any source fails. How it works Runs on a monthly schedule (or manually for testing) and loads configuration values like org IDs, lookback windows, and notification targets. Queries the GitHub API for organization members without 2FA and records a pass/fail (or “evidence_unavailable” with remediation details) for CC-6.1. Calls AWS IAM to generate and download the credential report, reads root MFA status, and evaluates IAM user MFA and access key age against the rotation threshold for CC-6.1 and CC-6.2. Pulls Google Workspace Admin audit logs for the configured lookback period and directory user 2-step verification enrollment to flag privileged actions (CC-7.2) and 2FA gaps (CC-6.1). Merges all findings into a single evidence set, appends each row to the “Evidence” tab in Google Sheets, and builds a per-control rollup with posture and top remediation items. Creates a Google Docs summary document, writes the compiled evidence summary into it, exports the document as a PDF from Google Drive, and emails it to the configured recipient via Gmail. Posts a warning message to Slack if any evidence source failed and had to emit “evidence_unavailable” rows. Setup Create a Google Sheet with a tab named “Evidence,” then add Google Sheets credentials in n8n and set the Sheet ID and Drive folder ID in the workflow configuration. Add a GitHub personal access token that belongs to an organization owner with the read:org scope, and set your GitHub organization name in the workflow configuration. Add AWS credentials with permissions iam:GenerateCredentialReport, iam:GetCredentialReport, and iam:GetAccountSummary, and confirm the region and key age threshold values. Add a Google OAuth2 credential with admin scopes (admin.reports.audit.readonly and admin.directory.user.readonly), and set your Google customer ID and lookback days. Add Gmail credentials and set the notification email address, and add Slack credentials and set the target channel ID for failure alerts. An n8n automation workflow template by Ali Abdullah.

By Tomasz Kaczyński

Quick overview For small service businesses: a two-page n8n form turns your Data Tables price list into an instant price estimate, emails it to the customer, saves it and alerts you on Telegram, so leads get a number in seconds instead of waiting for a reply. How it works Page 1 of an n8n form collects name, email and phone. The active services are loaded from the price_list Data Table and page 2 is built from them, so the dropdown always shows your current services. The customer chooses a service, quantity, an optional express option and notes. The estimate is calculated: unit price × quantity, two volume discount tiers, express surcharge, per-service minimum charge, then VAT. The estimate gets a number (for example EST-20261018-4821) and is saved in the price_estimates Data Table. You get a Telegram alert, the customer gets the estimate by Gmail, and the full breakdown is shown on the last form page. Setup Run the "Run once: create tables and sample prices" branch. It creates both tables and adds 4 sample services. Re-running it does not overwrite your prices. Replace the sample rows in price_list with your services, units, prices and minimum charges. In Settings, set business name, currency, VAT rate, discount tiers, express surcharge, validity days and Telegram chat ID. Add Telegram and Gmail credentials, activate the workflow and link the form from your website. Setup takes about 15 minutes. Requirements An n8n instance with Data Tables and multi-page Forms; a Telegram bot and chat ID; a Gmail account (or any email node). A Gmail credential (one click on n8n Cloud; on self-hosted n8n it needs a Google Cloud OAuth client), or replace the Gmail node with Send Email (SMTP) or Outlook. Customization Change tiers, surcharge and VAT in Settings; set active to false to hide a service. Replace Gmail with Outlook or SMTP and Telegram with Slack. Add fields such as address or preferred date on page 2, or turn accepted estimates into invoices or bookings. Additional info Volume discount tiers are global (they use the quantity of any service). For per-service tiers, add tier columns to price_list and adjust the Calculate the estimate node. The result is an estimate; the final price is confirmed by you. If the email cannot be sent, the customer still sees the estimate and is told that you will get in touch. If the form is public, link it only from your own website or add a captcha: each submission sends an email from your Gmail account. An n8n automation workflow template by Tomasz Kaczyński.

By AI Automation Playbook

Quick overview This workflow turns website visitors into qualified leads. Local service businesses — plumbers, clinics, salons, contractors — whose websites get visitors but few enquiries, and agencies that want to offer a chatbot service without building one from scratch. How it works A website visitor chats through a webhook endpoint. An AI agent answers questions from your FAQ knowledge base (Supabase vector store). An information extractor pulls the lead's name, phone, service needed and preferred time from the conversation. Qualified leads are saved to Google Sheets. The owner gets an instant SMS via Twilio. The full chat transcript is emailed to the owner via Gmail. Every Monday at 8:00 AM, the knowledge base updater lane downloads your FAQ document from Google Drive, re-chunks it, and upserts it into the same Supabase table the chat agent searches — answers stay current without manual work. Setup Connect your own OpenAI, Supabase, Google Sheets, Google Drive, Twilio and Gmail credentials (all slots are placeholders — none are included). Fill both CONFIG nodes (business name, trade, service area, hours, owner phone/email, leads sheet ID, Twilio from-number, Drive file ID of your FAQ document). Create a Supabase table named faq_documents and a Leads tab in your sheet. Upload your FAQ document to Google Drive. Run the weekly lane once manually to populate the knowledge base. Activate the workflow and point your website chat widget at the Chat Webhook production URL. Requirements An OpenAI API key, a Supabase project, a Google account (Sheets + Drive), a Twilio account with a phone number, and a Gmail account. A website chat widget that can POST to a webhook URL. Customization Edit the agent's system prompt to change tone and qualification rules, extend the information-extractor schema with extra fields (budget, postcode), swap the Sheets logger for your CRM, change the weekly schedule, or replace the Drive FAQ source with a Notion or website crawl. Additional info Useful for freelancers, small agencies, founders and SDRs doing their own cold outreach, plus agencies wanting a sellable chatbot service for local clients. An n8n automation workflow template by AI Automation Playbook.

By iamvaar

Quick overview Youtube Video: https://youtu.be/4eBS3peutkI This workflow captures Facebook Lead Ads submissions, validates and normalizes contact data, upserts the lead into GoHighLevel, and immediately sends a WhatsApp welcome template while saving the first message to Redis so a Google Gemini–powered WhatsApp agent can continue the conversation with context. How it works Triggers when a new lead is submitted via Facebook Lead Ads. Normalizes the payload (email, name, and phone), validates required fields, and stops processing for invalid leads. Searches GoHighLevel for an existing contact by email and creates/updates the contact when no match is found. Prepares a personalized WhatsApp outreach message, sends a WhatsApp approved template to the lead, and stores the same initial message in a Redis list keyed by the lead’s phone digits. Triggers on incoming WhatsApp messages and ignores any non-text messages. Uses a Google Gemini chat model with Redis chat memory to generate a short sales-assistant reply based on the inbound message and prior context. Sends the generated reply back to the lead via WhatsApp. Setup Connect Facebook Lead Ads OAuth2 credentials and select the correct Facebook Page and Lead Form in the Facebook Lead Ads trigger. Connect GoHighLevel (HighLevel) OAuth2 credentials and confirm the contact lookup uses the right identifier (email) for your CRM. Connect WhatsApp Business API credentials, set the correct Phone Number ID, and ensure the “lead_welcome” WhatsApp message template exists and is approved for your account. Connect Redis credentials and ensure your Redis instance is reachable for storing and retrieving chat history. Connect Google Gemini (Google PaLM) credentials for the LangChain chat model used by the WhatsApp AI agent. Update environment-specific defaults as needed, such as the default country code used to normalize 10-digit phone numbers in the validation logic. An n8n automation workflow template by iamvaar.

By Oneclick AI Squad

Quick overview This workflow reconciles incoming bank transactions against open invoices and existing ERP payment records, optionally escalating borderline matches to Anthropic Claude for review, persisting results in Postgres, and notifying a finance team via Slack while returning a run summary to webhook callers. How it works Starts a reconciliation run via a manual trigger, a daily 06:00 schedule, or a POST webhook request. Ensures the required Postgres tables exist and loads previously reconciled bank transaction and invoice IDs to make the run idempotent. Fetches open invoices and ERP payments from an ERP API and bank transactions from a bank API, then normalizes all three datasets into a consistent schema. Validates that all source API calls succeeded, then runs a rules-based matching engine to score and classify matches and generate exceptions (for example unmatched payments, overdue invoices, duplicates, and ERP conflicts). Sends borderline matches to Anthropic Claude for approve/reject/escalate decisions and applies those decisions with confidence and variance guardrails. Upserts matches, exceptions, and run summaries into Postgres, optionally posts auto-matched payments back to the ERP API with an idempotency key, and sends a Slack webhook alert only when items need review or exceptions exist. Returns the final reconciliation summary as JSON to the webhook caller or logs the run as failed in Postgres and responds with HTTP 502 if any source API fails. Setup Add a Postgres credential and ensure the database user can create tables and read/write the reconciliation_runs, reconciliation_matches, and reconciliation_exceptions tables. Update the source API URLs, query parameters, field mappings, and authorization headers for your ERP and bank systems in the configuration code. Add your Anthropic API key (and optionally disable AI review) by updating the Claude settings in the configuration code. (Optional) Create a Slack incoming webhook and set its URL in the configuration code to receive alerts when attention is needed. (Optional) Enable ERP write-back and set the ERP apply-payments endpoint and credentials after validating results in dry runs. An n8n automation workflow template by Oneclick AI Squad.

By Oneclick AI Squad

Quick overview This workflow runs on a schedule, via webhook, or manually to detect subscription revenue anomalies by pulling subscription events, payments, and optional business context from external APIs, analyzing daily metrics in Postgres, investigating likely causes (including an Anthropic Claude call), and alerting via Slack and PagerDuty. How it works Runs on a daily cron schedule, via a POST webhook (optionally with a targetDate override), or manually. Creates/ensures Postgres tables for metric history, anomaly records, and run tracking, then loads recent metric history and previously raised anomalies for baselines and deduplication. Fetches subscription events, payment records, and optional business events from external HTTP APIs, normalizes the data, and stops with a logged failed run if required billing feeds are unavailable. Builds daily metric series (overall and by plan/country/channel), applies data-quality gap checks, and detects statistically unusual point anomalies and sustained shifts using weekday-aware robust baselines. Investigates the most important anomalies by drilling down into contributing segments, correlating with business events, analyzing payment failure codes, and checking for single-customer concentration. For medium-or-higher investigated anomalies, sends evidence packs to Anthropic Claude to generate a root-cause narrative and recommended actions, with guard rails that only allow severity downgrades. Persists metric snapshots, anomalies (including evidence and AI verdicts), and the run summary to Postgres, then posts a digest of new anomalies to Slack and triggers PagerDuty for new critical unfavorable anomalies. Setup Add a Postgres credential and ensure the connected database user can create tables and read/write to revenue_metrics_daily, revenue_anomalies, and revenue_anomaly_runs. Update the feed URLs, auth headers/tokens, field mappings, FX rates, metric thresholds, and alert routing values in the workflow’s configuration code. Configure your Slack incoming webhook URL for alerts, and (optionally) a PagerDuty Events API v2 routing key for paging. Add an Anthropic API key (or disable AI in the configuration) if you want automated AI-written investigations. If using the webhook trigger, copy the webhook URL and call it from your system with an optional JSON body like {"targetDate":"YYYY-MM-DD"} for replays/backfills. An n8n automation workflow template by Oneclick AI Squad.

By Jesus Rojas

Quick overview Run your own video show with an AI host who looks the same every episode. Claude writes each episode, Shotstack puts the host in a new scene and makes them say the lines, lip-synced, with captions. Self-hosted n8n only: it saves files on disk. How it works Your show holds the show's name, topic, host, voice and camera style, the shots of each episode (one line per 10-second take) and how many episodes to make in a run. An n8n Data Table, made on the first run, logs every episode and keeps the host's picture, so the same host appears every time and episode numbers keep counting. With no picture yet, Draw the host makes one with Nano Banana 2, once. You can paste a link to your own character instead. Claude writes the next episodes: a new idea no earlier episode covered, a title, a scene, and for each shot about 20 words for the host to say and how the host moves, flowing as one talk. One Shotstack render per episode: Nano Banana 2 Edit puts the host in the scene, Gemini Omni Flash makes one lip-synced talking take per shot, the takes play back to back, and captions follow the speech, inside a fixed podcast-style layout. If an AI step fails, the render is retried once. Setup Install the Shotstack community node, version 0.2.1 or later, from the nodes panel or Settings → Community nodes, then import this workflow. Connect your Anthropic account in the Claude and Claude (fixes) steps, then connect your Shotstack account in Draw the host, Make the video and Check the render. No Shotstack account yet? Start a free trial: https://shotstack.io/?utm_source=n8n&utm_medium=template&utm_campaign=ai-show&utm_content=description Make an empty show folder in n8n's files folder (/home/node/.n8n-files in Docker). On an npm install, change that path in Save the episode. Fill in Your show with your show, host, camera style and shots, then click Execute workflow. To post every day, set episodes per run to 1 and turn on Every morning. Requirements Self-hosted n8n with Data tables, because the workflow saves files on disk The Shotstack community node, version 0.2.1 or later A Shotstack API key with credits for AI generation An Anthropic API key Customization Change what each shot is about, and how many takes an episode has, in Shots, in Your show Change the topic, host, voice and call to action in Your show Change the show's layout, colours and fonts in Build the episode Save the episodes to Google Drive or post them to YouTube instead of saving them on disk. An n8n automation workflow template by Jesus Rojas.

By Matt Michels

Watches the career pages of the companies you list, keeps only new postings, scores them against your target titles, level, location and pay, and routes them. Built for recruiters, job seekers and sales teams who use hiring as a buying signal. How it works Daily schedule / manual run reads every setting from Config. A weekly schedule uses the same Config to send a report instead of running the actor. Starts the Apify actor asynchronously, then waits and polls the run; FAILED, ABORTED, TIMED-OUT or too many polls stop with an error. Downloads the dataset, validates rows and dedupes within the run and across runs (workflow static data, dedupeMemoryDays). A transparent rule score runs first, then an LLM chain with a structured output parser reads each record against your plain-English aiCriteria and returns a fit score, verdict, summary, outreach angle and red flags. Calls are batched with a delay for rate limits, retried, and fall back to the rule score if the model fails. Upserts every row to Google Sheets. Hot rows go to Slack as one approve / dismiss message (send-and-wait, 24 h limit) and the decision is written back to the sheet; warm rows go to an HTML email digest; cold rows stay in the sheet. Weekly report: reads the sheet, summarises the last 7 days (tiers, approvals, top 10) and emails it. A separate Error Trigger branch posts any failure to Slack. Setup (about 15 minutes) Create an HTTP Header Auth credential: name Authorization, value Bearer (Apify Console > Settings > API & Integrations). Select it on the 3 Apify HTTP nodes. Create a Google Sheet with a tab named in sheetTab and a header row matching the sheet columns (see Upsert to Google Sheets, plus status and reviewedAt). Paste its ID into googleSheetId. Connect OpenAI (or swap the chat model sub-node for Anthropic, Gemini, Ollama...), Slack and Gmail; set slackChannelId and digestEmail in Config (also the channel on Slack: error alert). In Config set actorInput.companies (board slugs or career-page URLs), titleKeywords, preferredLocations, seniorityLevels, minSalary, hotScore, warmScore. Write your ideal-lead description in aiCriteria; set aiWeight (0 = rules only, 1 = AI only). Run once manually, check the sheet, then activate. Customize Edit weights in Score job relevance (title match 35, level 15, recency 15, remote 10, location 10, pay 15). Add keywords, departments, remoteOnly or postedWithinDays to actorInput to filter at the source. Swap Gmail/Slack for Teams, Discord, HubSpot, Airtable, etc. Each delivery node only reads $json. Disclosure: I maintain these Apify actors (account transparent_meteorite). The actor is a paid pay-per-event Apify tool; check its Store page for current pricing before scheduling. An n8n automation workflow template by Matt Michels.

By hqasmei

An AI-powered application hosted on Hugging Face Spaces by hqasmei. Built with gradio.

By Cierpliwy

Paint inside your line art with Krea 2 Hosted on Hugging Face Spaces by Cierpliwy. SDK: gradio.

By melodydeyoung

An AI-powered application hosted on Hugging Face Spaces by melodydeyoung. Built with docker.

By hugging-apps

Feedforward 3D Gaussian reconstruction from driving views Hosted on Hugging Face Spaces by hugging-apps. SDK: gradio.

By VisionXLab

Official Leaderboard of RISEBench Hosted on Hugging Face Spaces by VisionXLab. SDK: static.

By VecSzn

35B MoE models on an 8 GB GPU: speeds and install Hosted on Hugging Face Spaces by VecSzn. SDK: static.

By earakely-scale

50 replay environments REA built from app sessions Hosted on Hugging Face Spaces by earakely-scale. SDK: static.

By lucas1026

Source-linked evidence of evolving AI agent harnesses Hosted on Hugging Face Spaces by lucas1026. SDK: static.

By hugging-apps

Detect, ground, count, OCR, segment & matte with one VLM Hosted on Hugging Face Spaces by hugging-apps. SDK: gradio.

By FlameF0X

An AI-powered application hosted on Hugging Face Spaces by FlameF0X. Built with gradio.

By Saiyejin

An AI-powered application hosted on Hugging Face Spaces by Saiyejin. Built with docker.

By lutetjeff

An AI-powered application hosted on Hugging Face Spaces by lutetjeff. Built with static.

By hugging-apps

Play chess against a tiny reasoning LLM that thinks first Hosted on Hugging Face Spaces by hugging-apps. SDK: gradio.

By michaldobiezynski

An AI-powered application hosted on Hugging Face Spaces by michaldobiezynski. Built with static.

By harvis-io

Describe a site, any model writes it, ship it to harvis.dev Hosted on Hugging Face Spaces by harvis-io. SDK: static.

By k1ngtai

Burmese Text to Speech using Gemini Hosted on Hugging Face Spaces by k1ngtai. SDK: gradio.

By 0xmoose

An AI-powered application hosted on Hugging Face Spaces by 0xmoose. Built with docker.

By RUC-NLPIR

An AI-powered application hosted on Hugging Face Spaces by RUC-NLPIR. Built with gradio.

By ZiruiWu

An AI-powered application hosted on Hugging Face Spaces by ZiruiWu. Built with gradio.

By Ntdeseb

Modelos libres de IA optimizados para ZeroGPU H200 Hosted on Hugging Face Spaces by Ntdeseb. SDK: gradio.

By Kukulauren

3D imaging Hosted on Hugging Face Spaces by Kukulauren. SDK: streamlit.

By pinyuchen

An AI-powered application hosted on Hugging Face Spaces by pinyuchen. Built with gradio.

By SeyedAli

An AI-powered application hosted on Hugging Face Spaces by SeyedAli. Built with gradio.

By SlayerLab

Polish 1B LM trained from scratch, version by version Hosted on Hugging Face Spaces by SlayerLab. SDK: static.

By gchauhan

An AI-powered application hosted on Hugging Face Spaces by gchauhan. Built with static.

By Osman-Ozcanli

Streamlit template space Hosted on Hugging Face Spaces by Osman-Ozcanli. SDK: streamlit.

By jjiantong

System-Aware KV Cache Optimization for LLM Serving [Survey] Hosted on Hugging Face Spaces by jjiantong. SDK: static.

By Kwasiasomani

An AI-powered application hosted on Hugging Face Spaces by Kwasiasomani. Built with docker.

By nebius-build-week

Find the papers your search missed, on a shared board Hosted on Hugging Face Spaces by nebius-build-week. SDK: static.

By xyz.clickthislink

Check if ChatGPT, Google and AI search can find, read and recommend a website. Free, no key.

By xyz.tenzro

Canton ledger tools over MCP: DAML commands, CIP-56 assets, DvP settlement, party management.

By xyz.tenzro

Chainlink tools over MCP: CCIP cross-chain, data feeds, data streams, VRF, proof of reserve.

By xyz.tenzro

Ethereum tools over MCP: Chainlink feeds, ENS, ERC-8004 agents, EAS attestations, transactions.

By xyz.tenzro

Solana tools over MCP: Jupiter swaps, SPL tokens, Metaplex NFTs, SNS domains, network stats.

By xyz.urbangametheory

Inspect real parcels, create and edit urban plans, compare alternatives, and share 2D/3D designs.

By tech.tradetron.mcp

Build, backtest and diagnose algo trading strategies in your own Tradetron account in plain English.