Start free. Scale when it works.

Prospecting is the most expensive agent to run: it harvests live buying signals, then googles each person to locate their real LinkedIn profile and match their identity against company and role context. That is several extra searches and model calls per prospect. Starter gives you the full community-led motion (finding threads and drafting replies), while Growth adds named buyers, their verified profiles, and the outreach written for each one.

Hallucination is engineered out with code-level gates the model cannot talk past. Every URL must appear verbatim in the fetched source material, every evidence quote must literally exist in the original post, a second skeptical model re-reviews each candidate against your product, tweets are checked against the live oEmbed endpoint, and emails are only kept when they appear in the sources (otherwise they are labeled unverified). Anything that fails a single gate is dropped, and an empty cycle is treated as a correct answer rather than padded with guesses.

A Mission is a hard target you hand the agent team, like "find my next 100 customers." Progress is computed from real workspace data and the plan re-evaluates every few hours until the number is hit. Missions are on Custom plans only right now: they steer other agents and spawn campaigns on their own, so we are running them hands-on with each team until that loop has more mileage behind it.

It sends email when you press send, and it never posts. On Growth you connect your mailbox over SMTP, open a drafted email, edit it, and click send: it goes immediately from your own address, up to 500 a month. Nothing is scheduled and nothing is queued, so an email only exists because you read it and sent it. Replies, DMs and social posts are drafted for you to publish yourself, because your account standing in a community is yours and no forum should ever see automated traffic arriving from us. Drafting is never metered on any plan.

No. Frontier-model usage is included in your subscription, there are no separate token bills and no API keys to set up. Revtive runs every agent and chat for you and routes each task to an appropriate model automatically.

Yes. Revtive maps your ideal customer first, then plans the venues those people actually post in: broker boards for real estate, operator communities for restaurants, practice-management communities for clinics, GitHub and Stack Overflow for developer tools, LinkedIn and procurement boards for agencies. General web search runs alongside the vertical venues on every plan that includes venue discovery.

It cannot. There is no scheduling or publishing path in the build at all: nothing is queued and nothing goes out on a timer, so no agent can reach another person on its own. The only way a message leaves is a person opening the draft and pressing send, and that is Growth only. On top of that, each agent can be tuned or switched off on its own, integrations are gated behind admin feature flags, and one master switch pauses the whole team instantly.

Yes, on the Custom plan. Beyond the built-in agents, you can define your own with their own prompt, schedule, and capabilities, invite users and teams into the workspace, and set per-user permissions for who can run what.

Most workspaces see their first qualified, evidence-backed opportunities within 24 hours of adding a URL. The system compounds from there: agents learn from every accept, dismiss, and edit you make.

Yes. Plans are month-to-month, downgrades take effect at the end of the billing period, and your workspace data remains exportable.