AI Agent · ChatGPT
A ChatGPT chatbot that only answers from your content
ChatFlo runs your support agent on OpenAI's GPT-4o models, grounded in your own website, docs, and policies — so it answers like someone who works at your company, not like a general-purpose chatbot.
Can ChatFlo run my support agent on ChatGPT?
ChatFlo runs your support agent on OpenAI's GPT-4o models and grounds every reply in your own website, documents, and policies. You do not need an OpenAI account or API key — model usage is included in your plan. Answers cite their sources, and anything the agent cannot resolve is escalated to your team with the full thread.
Why ChatGPT
The default, and the one most teams should start on.
The most predictable starting point
GPT-4o is ChatFlo's default because it handles the widest range of support phrasing without tuning. If you have no strong opinion about models, this is the one to leave selected.
Strong at messy, real-world questions
Customers rarely ask cleanly. GPT-4o holds up well on run-on messages, several questions at once, and typo-heavy mobile writing.
A cheaper sibling for high volume
GPT-4o mini handles routine, high-volume traffic at a fraction of the cost, so you can run a smaller model on FAQs and reserve the larger one for complex threads.
Available models
The OpenAI models you can select
Pick one in AI settings. Switching is a setting, not a migration — your training content, guardrails, and history stay where they are.
GPT-4o
Smart, and ChatFlo's default
GPT-4o mini
Fast and cheap for high volume
GPT-4.1
Stronger reasoning on complex threads
GPT-4.1 mini
Balanced cost and capability
When another model is the better call
If your priority is long, carefully-worded replies that closely follow a written policy document, Claude tends to be the better fit. If you need very large documents handled in a single pass, look at Gemini.
Put GPT-4o on your support queue
Point ChatFlo at your website, and your agent starts answering from your own content in minutes.
Frequently asked questions
A custom GPT lives inside ChatGPT, where your customers are not. ChatFlo puts the same class of model on your website, WhatsApp, Instagram, Messenger, SMS, and LINE, with a shared inbox behind it so your team can take over a conversation mid-thread. You also get source citations, escalation rules, and per-channel analytics that a custom GPT has no concept of.
The agent answers from the material you gave it and cites the sources behind each reply, so any claim can be checked in a tap. When it has no confident, grounded answer it hands off to a human rather than filling the gap.
No. ChatFlo runs the model for you on every paid plan — there is no separate OpenAI account, billing relationship, or key to manage. Model usage is included in your plan.
Yes, from the AI settings screen, at any time. The model is a setting on your agent, not a plan tier — your training content, guardrails, tone, and conversation history are untouched by the switch, so you can A/B two models on the same setup.
No. Training is the same regardless of model: you point ChatFlo at your website, docs, and files, and it indexes them. The model only affects how answers are composed from that material — it never changes what the agent is allowed to draw on.
Other models
ChatFlo is not locked to one model vendor
Your agent runs on whichever of these you select, and you can change your mind without rebuilding anything.