Assistant web research: cheaper, and now itemised
When you ask the assistant to research a company or read a page, it now runs on a faster, far cheaper model — the tokens for a research call now cost a small fraction of what they did, roughly a twentieth. The web searches those calls make are now billed as their own line, at 1.5 cents per search. Even with the searches itemised, a typical research call totals about a fifth of what it did before (around 6 cents, previously around 25), so this applies immediately.
- Assistant web research: 1.5 cents per search (previously not billed separately), plus the conversation's own token metering — which for a research call is now roughly a twentieth of what it was.
- Why the searches are itemised: a research call pays a per-search fee on top of tokens. It was absorbed before; showing it is what lets the token side come down.
- Nothing else about the assistant changes: reading, planning and building campaigns in the conversation are metered exactly as before, and the assistant is still deliberately the cheapest thing you can do in Boostie.
- Prospective only: conversations already billed are never re-billed.
- The pricing page and the assistant's own quotes use the new rate from the same moment, so what you are quoted is what you are billed.
- No other price changes: leads, verification, scoring, AI writing, sending, inboxes and domains are unchanged.