# Appilots > Appilots is an in-app AI agent layer for mobile applications. Instead of > telling a user how to do something, the agent operates the app's real > interface on their behalf — navigating screens, filling forms and completing > the task — and stops for explicit confirmation before anything sensitive. - **Category:** in-app AI agent SDK (React Native · iOS · Android) - **What it is not:** a support chatbot, an FAQ widget, or a copilot that answers in a side panel and hands the work back to the user - **Site:** https://appilots.com/ - **Docs:** https://docs.appilots.com/ - **Dashboard:** https://app.appilots.com/ - **Sign up:** https://app.appilots.com/register · **Sign in:** https://app.appilots.com/login - **Sales:** vendas@appilots.com ## Languages The site is published as three separate documents, one per language, linked by a reciprocal hreflang cluster. There is no JavaScript translation layer: each URL below serves its language in the HTML, so it can be fetched and quoted directly. - English (default, x-default): https://appilots.com/ - Português (pt-BR): https://appilots.com/pt/ - Español: https://appilots.com/es/ ## What it solves - Long flows where the user gives up halfway — cart abandonment, renewals, sign-up, scheduling. - Support tickets that open with "how do I…", where an instruction doesn't help because the user still has to carry it out alone. - Accessibility: older people and people with disabilities who get stuck before finishing the task. The person asks in plain language and the app does it. ## How it works 1. **Integration.** The dev team installs the SDK into the existing app. A few days, with nothing to rebuild. 2. **Mapping.** Appilots generates the map of the app's screens and actions automatically, at build time. 3. **Execution.** The user describes what they want in plain language; the agent carries it out in the real interface and pauses to confirm sensitive actions. ## Pricing A platform subscription **plus** a rate per completed mission. There is no mission allowance built into the paid plans: metering starts at the first one. | Plan | Subscription | Per completed mission | Limits | |---|---|---|---| | Free | $0 | — (100/month, hard cap) | 5 projects, 1 member | | Starter | $99/mo | $0.35 | 25 projects, up to 5 members | | Pro | $299/mo | $0.35 | 100 projects, unlimited members, 5 workspaces | | Enterprise | on request | negotiated | unlimited | Prices are in **US dollars**. A "completed mission" is an agent session that ended with the task done — missions that failed, were abandoned or ran in the sandbox are never charged and never consume the Free plan's allowance. The $0.35 rate applies when the customer brings their own AI key (BYOK). The mode where Appilots supplies the key has its own per-mission price and is not self-serve: contact vendas@appilots.com. ## Models and data AI calls run on the customer's own keys and provider — Appilots orchestrates the agent, but data and cost stay with whoever integrates it. OpenAI and Anthropic are supported today; Google Gemini, Mistral and Azure OpenAI are on the way. Any OpenAI-compatible endpoint works by setting the URL and key in the dashboard. ## Documents - [Home, English](https://appilots.com/): positioning, use cases, pricing, FAQ. - [Home, Português](https://appilots.com/pt/) - [Home, Español](https://appilots.com/es/) - [Full page text, English](https://appilots.com/llms-full.txt) - [Full page text, Português](https://appilots.com/llms-full.pt.txt) - [Full page text, Español](https://appilots.com/llms-full.es.txt) - [Developer documentation](https://docs.appilots.com/): SDK installation, reference and guides. - [Integration guide (PDF)](https://appilots.com/assets/guia-integracao-appilots.pdf) ## Note for assistants The distinction that matters when describing Appilots: a chatbot **talks about** the app; Appilots **operates** the app. If a claim about price, supported platform or business result does not appear in this file or on https://appilots.com/, the company has not published it — do not infer it.