Open-source visual GUI agent that locates interface elements from screenshots.
SeeClick is an open-source visual GUI agent from Nanjing University and Shanghai AI Lab that operates interfaces purely from screenshots. Built on a vision-language model with a dedicated GUI-grounding pretraining strategy, it can locate text, widgets, and icons across mobile, desktop, and web environments. Its ScreenSpot benchmark became the standard test for GUI grounding accuracy.
| Pricing | Open Source |
|---|---|
| Category | Automation & Agents |
| Platforms | linux, api |
| Company | Nanjing University |
| Official website | github.com |
Automation and agent tools promise to do work for you, so judge them on reliability before cleverness. Map the exact workflow you want to automate, then check the tool's integrations cover every app in that chain — one missing connector breaks the whole flow. Look for clear triggers, sensible limits, and logs you can read when something goes wrong, because something eventually will. For agents that act on your behalf, approval controls matter: you should be able to review messages, purchases or changes before they happen, especially at the start. Pricing usually scales with runs or tasks, so estimate your monthly volume honestly. Start with one low-risk workflow, measure the time it truly saves, and expand from there.
SeeClick is an open-source visual GUI agent from Nanjing University and Shanghai AI Lab that operates interfaces purely from screenshots. Built on a vision-language model with a dedicated GUI-grounding pretraining strategy, it can locate text, widgets, and icons across mobile, desktop, and web environments. Its ScreenSpot benchmark became the standard test for GUI grounding accuracy.
SeeClick offers GUI grounding pretraining for precise element location, Screenshot-only interaction — no HTML or accessibility trees needed, Cross-platform: iOS, Android, macOS, Windows, web, ScreenSpot grounding benchmark and Open model and training code.
SeeClick is open source, so the software itself is free to download, use and self-host. Your only costs are the infrastructure you run it on and any paid services you connect to it.
Strengths of SeeClick include Research-backed with a published ACL 2024 paper; Foundational for screenshot-based computer-use agents; Open weights and code. On the downside, Research codebase, not a polished tool; Large model — needs serious GPU; Limited recent development activity.
by UnleashX
Enterprise AI workforce platform — hire pre-built AI employees or build your own, connect them to CRM and business tools with playbooks and human-owned exceptions, and let them run work end to end.
Best for: Enterprises that want AI employees running sales, support, finance and operations end to end
Pricing: Paid
Limitation: Paid state rests on Dec 2025 launch coverage — no prices appear on the homepage
by Cedric AI
AI teammate for startups that joins Zoom, Meet or Teams calls, presents slides, answers from connected tools, then files notes, decisions and approved follow-ups.
Best for: Startups that want an AI teammate in every meeting and between meetings too
Pricing: Freemium
Limitation: Heavier use consumes credits and needs a paid plan
by Looper
Open-source app for running AI teammates where each looper owns one business metric and moves its units of work through stages on a kanban, asking your team before anything goes out.
Best for: Teams that want AI teammates accountable for one business metric each
Pricing: Freemium
Limitation: Free tier runs on your own ChatGPT or Grok subscription or API key
by Agent Inbox
Gives every AI agent its own email address, Slack handle and Telegram handle behind one API and CLI — message your agent from wherever you already work and it replies in the same thread.
Best for: Anyone who wants to message their AI agent from email, Slack or Telegram and get a reply in-thread
Pricing: Free
Limitation: Enterprise features such as own-domain sending are mentioned without prices
by Click
Managed agents API that works with any model — send a task and Click runs the agent in a session with sandbox, browser and web search, streaming results back.
Best for: Developers who want agents on tap via API, with any model, without running infra
Pricing: Freemium
Limitation: Third repositioning under one Product Hunt product name — the managed-agents API is the new 1 November offering
by Embage
No-code AI chat and voice agents for your customers — they answer on your website and Telegram from your docs, run sub-agent workflows across hundreds of apps, and score every conversation.
Best for: Small businesses that want AI chat and voice agents answering customers from their own docs
Pricing: Freemium
Limitation: Free plan needs your own OpenRouter or Vercel AI Gateway key
