AI chat with web search: every Whizi model answers from live sources

In brief

Whizi is an AI chat with web search on every model, so the model you selected answers from current sources instead of only from its training data. It costs no credits and works with whichever model you pick. There is no switch to set: Whizi checks each message and runs a real search only on the ones that need it.

The short answer

Every model in Whizi can search the web, and the search is automatic. The model you selected answers from current sources instead of only from its training data, it costs no credits, and the decision is made per message rather than per model, so it applies to whichever model you pick.

Two checks decide whether a message gets searched. First, a small, fast classifier reads your message before the model does and judges whether it needs the web. When it says yes, the search runs straight away. When it says no or is unsure, the model still holds a cheap search function and can call it if it decides it needs a source. A request for a poem gets no search, and a question about this week does.

The per-message part is the useful bit. A model with no retrieval of its own normally answers a question about this month from a training set that ended months ago. Whizi puts a search layer in front of whichever model you picked, which means you can ask Claude, or a Llama model, or a small fast model, a question about today and get a sourced answer.

If an answer you expected to be sourced came back without sources, or the message errored, the failure surface is in web search not working.

When it changes the answer

Three categories where the difference between a searched answer and an unsearched one is not subtle.

Anything with a date in it. Prices, model releases, product launches, company news, version numbers. Without search, a model answers confidently from whatever was true when it was trained, and gives you no signal that the information is stale. This is the single most common way AI answers go quietly wrong.

Anything obscure or specific. A small company, a niche tool, a local regulation, a particular error message. Training data covers these thinly or not at all, and thin coverage is exactly the condition under which models invent plausible detail.

Anything you intend to act on. Search produces sources you can open. A claim with a link you can check is a different object from a claim without one, even when both turn out to be correct.

When to keep the web out

Search is not free of cost to answer quality, which is why Whizi skips it on messages that do not need it.

For reasoning, writing and analysis of something you supplied, search adds nothing and can actively distract. If you have uploaded a document and want the model to reason about that document, pulling in web results introduces material that is not yours and was not asked about. Same for creative writing, code review of your own code, and working through a problem step by step.

The rough rule: search helps when the answer depends on facts about the world, and gets in the way when the answer depends on thinking about something already in the conversation. To keep it out, say so in the message, for example "answer only from the document I uploaded, do not search".

What you are asking aboutSearch helpsWhy
Prices, releases, launches, company news, version numbersYesWithout it the model answers from training data and gives no signal that the information is stale
A small company, a niche tool, a local regulation, a particular error messageYesTraining data covers these thinly, and thin coverage is when models invent plausible detail
Anything you intend to act onYesIt produces sources you can open, which is a different object from a claim without one
Reasoning about a document you uploadedNoRetrieved pages introduce material that is not yours and was not asked about
Creative writing, or code review of your own codeNoThe answer depends on thinking about something already in the conversation

Reading the sources properly

A sourced answer is more checkable, not automatically more correct. Two failure modes survive retrieval and are worth watching for.

The first is a source that exists but does not support the claim. The model retrieved a page, and the page is real, but the specific number or statement attributed to it is not in it. Opening one or two cited links on any answer you plan to rely on catches this quickly.

The second is a source that is itself wrong or out of date. Search does not evaluate credibility, so a stale blog post ranks like an authoritative one. For prices and specifications especially, a vendor page beats a roundup article, and both beat a forum comment.

If a claim matters, the fastest verification is to ask a second model the same question, in the same conversation, and say you want current sources. Where two independent retrievals agree on a specific number and a source, confidence is reasonable. Where they diverge, that is your finding.

Workflow checklist
  • Search is automatic: there is no search button in the composer
  • Web search costs no credits
  • Search is decided per message, not per model, so it works with whichever model you pick
  • Whizi searches only on the questions that need it, not on every message
  • Say "do not search" in the message when you want an answer from your document alone
  • Open at least one cited link on any answer that matters
  • Cross-check important claims by asking a second model for current sources
Common questions

Frequently asked questions

Can Claude search the web in Whizi?

Yes. Web search applies to whichever model you selected, so Claude Sonnet 5 or Claude Opus 5 answers from current sources. This is one of the practical advantages of a multi-model workspace: you are not limited to whichever models happen to ship with their own retrieval.

Does web search use extra credits?

No. A message costs the same whether it was searched or not, because the credit rate is set by the model alone. The only thing that changes the cost of a message is which model answers it.

Why did the model give me outdated information?

Both checks judged the question answerable without the web, so the model answered from training data. Ask again and say you want it looked up against current sources, which is a clear signal for a search. If the answer changes, the first one was stale.

Can I turn web search off?

There is no switch, because Whizi only searches when a message needs the web, so a message that does not need it costs nothing extra. When you want an answer drawn only from material already in the chat, such as a document you uploaded, say so in the message. In the mobile app the old Search button is still visible for now, but it no longer changes anything.