ChatGPT is the product that made this category, and its models are genuinely excellent. Two things it is not: it is one vendor’s models, and it is a chat window — your inbox, your week and your notes are somewhere else, and getting them in is your job. HAL is several vendors’ models sitting inside the place where those things already are.
| ChatGPT | HAL | |
|---|---|---|
| Models | OpenAI’s, and they are very good | Several vendors in tiers — you are not betting your whole assistant on one lab |
| Quality on a hard prompt | Frontier. Genuinely hard to beat | The deep tier is competitive; the cheap tier is for the questions that do not need it |
| Your inbox | Not part of it | An address at gethal.com that HAL reads, searches and replies from |
| Your calendar | Paste it in | Part of the product — HAL creates and moves events and reminders |
| Memory | Memory across chats, improving fast | One memory shared with mail, calendar, notes and the archive |
| Ecosystem | Huge: custom GPTs, plugins, a large developer community | Much smaller. This is a real gap and it is not close |
| A real human | No | Book a specialist from inside the app |
Written from each product’s public documentation. These are good, mature tools and this page says where they are still better. No prices here: they change on their own and the page would end up lying without anyone touching it.
If what you want is an assistant that can also see your mail, your week and your notes, yes. If what you want is the strongest single chat model with the largest ecosystem around it, ChatGPT is the honest answer.
On the model itself, that question does not have a stable answer — the labs pass each other every few months. HAL’s bet is not "a better brain", it is "the same class of brain with your context already in the room".
Because they are good at different things and they fail at different times. A tier system also means a one-line question does not get billed at the price of a hard one.
Not on its own. HAL gives you an address it owns end to end, so reading, searching and replying are part of the product rather than a connector you maintain.
It has a marketplace of small apps and agents you can use or build, and historical figures you can talk to. It is far smaller than the GPT store and pretending otherwise would be silly.
Plenty of people do, and nothing here argues against it.
A larger ecosystem, more third-party integrations, and a much bigger community writing prompts and tools for it. Its video also comes back in the same session, where HAL delivers a clip to your Archive minutes later.