Model fit
Choosing a model for research with sources
How ClawAI’s Research mode grounds an answer in sources it looked up, how that is billed separately from model credit, and what still depends on the model you choose. Confirm the live catalog before choosing a plan.
All tasks · Last reviewed:
What ClawAI’s Research mode does
Research mode lets a request search the web, fetch a page, or fetch and extract structured content from one, before the model produces an answer — so the answer can cite sources it retrieved for that specific question rather than relying only on what the underlying model learned during training. It is a plan-gated feature with three depths: search only, search plus fetch, or search plus fetch and extract.
Research is billed separately, not from your model credit
Research access is metered as its own usage — web search, page fetches and extraction — separate from the token allowance a chat message draws on. A claim that “research uses your model credit” would be wrong: the two are tracked and billed as different things, and your plan’s research allowance is a separate line from its model token allowance.
What the underlying model still changes
Research mode changes what the model can see before it answers, not how well it reasons over what it retrieved — a model still has to read the fetched sources, weigh them against each other, and write an answer that reflects them accurately. The same considerations that apply to complex reasoning tasks apply here: a model built for working through several steps is a reasonable fit for reconciling several sources, and that is worth checking against your own material rather than assuming.
Model availability and allowances are enforced by your plan and the live catalog, not by this page. Confirm the live catalog on the pricing page before choosing a plan built around a specific model. Confirm the live catalog on the pricing page
Questions people ask
- Does research use my model credit?
- No. Research access — web search, page fetches and extraction — is metered separately from the token allowance a chat message draws on. Confirm both allowances on the pricing page.
- What is the difference between the three Research mode depths?
- Search only returns results from a web search; search plus fetch also retrieves the page content; search plus fetch and extract additionally pulls structured content out of what it fetched. Which one a request uses depends on how it is configured.
- Does the model I choose matter if Research mode is on?
- Yes — Research mode changes what sources the model can see, not how well it reads and reconciles them. See how to evaluate AI models, linked below, for how to check that against your own workload.
Try it rather than take our word for it
ClawAI’s Research mode can search, fetch, and fetch-and-extract from the web before a model answers — a real, shipped feature, metered separately from your model token credit.