Frequently asked questions

Trainly lets you build an AI support agent for your website and train it until it answers correctly. You teach it from your own site and documents, test it like a real customer, correct anything it gets wrong in plain English, and embed it on any site with a single line of code.

No. You train the AI by chatting with it and clicking, and you deploy it by pasting one line of code. If you can write an email, you can use Trainly.

Sign in with your email or Google account and you can start straight away. There is no waiting list and no card required.

Start from the homepage: enter your website address and, optionally, a line about what you want it to help with (support questions, lead capture, explaining your services, product recommendations). Trainly reads your site and hands you into the builder to train and test it.

In the Training Studio, open the + menu and choose Add website or Upload files. On the Knowledge page you can also paste in plain text or add question-and-answer pairs directly. Everything you add becomes part of your agent’s knowledge, the single pool it answers from.

PDF, TXT, MD, PNG, JPG/JPEG, and WEBP. Images and PDFs are read with vision so text inside them is understood too.

Point Trainly at your site and it reads it top to bottom, in the background, a batch of pages at a time, until your site is in knowledge. It keeps going even if you move to another tab, and it reads safely, only fetching public pages with sensible page, time, and size limits.

It does a fast first pass and then continues in the background. In practice the limit is your plan’s knowledge allowance rather than a page count, and there’s a large safety ceiling so it never runs away on an enormous site.

Yes. Alongside websites and files, you can paste in text or add specific question-and-answer pairs, handy for policies, hours, or anything not written down on your site yet.

Yes. You can turn any source off. It stays in your library but is withheld from the AI, so it won’t be used in answers until you switch it back on.

Always. The Training Studio’s Test your AI panel lets you chat with your agent exactly like a customer would, so you see every answer before it’s ever published.

Flag the answer and describe the fix in plain English in the Train while you test panel. Trainly saves it as a rule, a fact, or an approved reply and applies it the moment you retest. No rewriting prompts, no code.

It has Trainly play a realistic customer and ask your agent questions, one message at a time, so you can stress-test it and surface gaps without waiting for real visitors.

It stays honest: it tells the customer it isn’t sure instead of guessing. Those moments are flagged so you can teach it the answer.

Paste one <script> line just before your closing </body> tag. It works with any website and needs no rebuild. The exact snippet, with your agent’s id, is in the Deploy tab (see the Docs for step-by-step instructions).

Chat bubble is a floating launcher that opens a compact popup. Sidebar is a full-height assistant drawer. You choose either one in the Deploy tab and can switch anytime.

Yes: logo, brand color, background style and image, assistant name, subtitle, greeting message, font, size, and corner radius are all yours to set in the Deploy designer, with a live preview.

On paid plans, yes. The Free plan keeps Trainly branding; Starter and Growth let you hide it.

No. The widget renders inside its own shadow root, so your site’s CSS can’t leak into it and its styles can’t leak into your page.

In the Chats tab. Every deployed conversation is saved, with thumbs-up/down feedback on individual answers and an outcome for each chat (resolved, or the customer left).

Yes. Every conversation your deployed widget has is saved in the Chats tab, on every plan.

Yes. Real conversations surface the gaps: liked and disliked answers show you what’s working and what isn’t, and you can turn any weak answer into new knowledge in a click.

Yes. It automatically detects the customer’s language and replies in it, falling back to English when it’s unsure.

Free is $0, Starter is $29/month, and Growth is $79/month. Plan changes are handled by us, so contact us to upgrade. See the full breakdown in the Docs.

Credits are a monthly allowance, shared across all the agents in a workspace, that gets used whenever your AI does work. That means answering visitors on your site, but also everything you do while building it: testing an agent, running a simulated customer, reading and indexing a website or document, and the suggested tests Trainly regenerates after you save new training. How much a single action costs varies with the model, the amount of context, and the response length, so credits aren’t a fixed number of messages. You can see what’s left at any time in the sidebar, and a full per-action breakdown on the Usage page.

No. Credits reset at the start of each monthly billing cycle — the day of the month your workspace started; unused credits don’t carry over.

Free includes 1 agent, Starter 3, and Growth 10.

The combined knowledge limit per workspace is 50 MB on Free, 0.5 GB on Starter, and 2 GB on Growth.

Yes, on paid plans. Starter includes 1 teammate seat and Growth includes 10 (the owner doesn’t count as a seat). You control which agents each teammate can open.

Yes. Plans belong to workspaces, not people, so a single account can own several workspaces on different plans.

Answers are generated with OpenAI’s GPT-4.1-mini by default.

Trainly stores the knowledge you give each agent and the conversations your deployed widget has, so you can review and improve them. For the full details on what’s collected and how it’s handled, see our Privacy Policy.