Viktor Pricing in 2026: What Credits Actually Cost
Viktor’s pricing page, checked on September 8, 2026, answers the first question quickly and the second one not at all. The first question is what a plan costs, and the page is clear. It’s $50 a month to start, and $100 for the tier it marks most popular. The second question is what you will actually pay, and that depends on how hard your team runs it, because every plan is a credit allowance rather than a license.
Viktor charges per workspace, not per seat
This is the fact that changes every comparison. Viktor’s Team plan is described as “One AI employee your whole team shares,” and credits are “shared across your whole workspace.” Their own FAQ puts it plainly: “Is there a per-seat charge? No. Viktor pricing is workspace-based. Your team shares the same workspace credits.”
So headcount doesn’t move the bill. A team of three and a team of thirty pay the same for the same credit allowance. Against per-seat tools, that comparison gets more favorable to Viktor with every person you add, and it’s worth doing carefully. A $50 workspace across ten people is $5 a head, which no per-seat product will match.
Usage moves the bill instead.
The full credit ladder
The Team plan is one dropdown with 26 rungs. Every rung is the same rate, $2.50 per 1,000 credits.
| Credits / month | Price | Credits / month | Price | |
|---|---|---|---|---|
| 20,000 | $50 | 400,000 | $1,000 | |
| 30,000 | $75 | 1,200,000 | $3,000 | |
| 40,000 | $100 (most popular) | 2,000,000 | $5,000 | |
| 80,000 | $200 | 5,000,000 | $12,500 | |
| 125,000 | $300 | 10,000,000 | $25,000 | |
| 200,000 | $500 | 20,000,000 | $50,000 |
Above and below those rungs the pattern holds all the way out to 20 million credits at $50,000 a month.
There is no volume discount
Look down the right-hand column and the rate never moves. Buying a thousand times more credits costs a thousand times more. The smallest rung and the largest are both $2.50 per 1,000.
A flat rate across a 1,000x range is unusual, and Viktor built it that way on purpose. Their Enterprise card lists “volume credit pricing with discounts” as one of the things Enterprise adds. The discount exists; you just have to book a call to get it. If you’re heading past a few thousand dollars a month on self-serve, that call is the whole negotiation.
What a credit buys
Viktor publishes its own guidance, and it’s the most useful part of the page:
- A quick Slack question: 50 to 100 credits
- A deep research report or multi-step automation: 500 to 2,000 credits
- Their summary: “Roughly 40-200 tasks depending on complexity. Most solo users find 20K-40K plenty for a month.”
Take the middle of that. At 40,000 credits, the $100 tier is around 400 quick questions, or roughly 30 research reports, or some mix. For one person that’s generous. For a team of ten sharing one allowance, 40,000 credits is 4,000 credits each, which is two research reports apiece per month.
That arithmetic is why the workspace model cuts both ways. Nobody pays for a seat, but nobody gets their own allowance either.
Rollover is one month, not indefinite
Their FAQ: “Unused monthly plan credits now roll over for one month.” Free trial and bonus credits are the exception and never expire.
Two things follow. A quiet month doesn’t bank credits you can spend next quarter, so buying a large rung against future need wastes most of it. And the widely repeated claim that Viktor credits expire monthly is wrong, but so is treating the balance as money in the bank.
Why reported bills run above the plan
Viktor ships three separate spending controls: top-ups on demand “when a big project hits,” auto-topups with a monthly spend cap, and a per-user spend cap reserved for Enterprise. Nobody builds three guardrails for a product whose customers sit comfortably inside the bundled allowance. The controls are the tell.
User reports line up with that. A team of five described exhausting 20,000 credits in 36 hours, a pace that works out near $1,000 a month. One user burned 40,000 credits during the trial and estimated $300 to $400 a month for their real workload, three to four times the headline tier. A G2 reviewer put it as “Viktor is not cheap, and in return it doesn’t give cheap work.”
Treat the direction as established and the exact figures as indicative. Much of this circulates through vendors who sell Viktor alternatives and have an interest in it reading as expensive, so the user-generated reports on G2 and Reddit carry more weight than the roundups. What none of them dispute is the shape. The plan you pick is a floor.
What to work out before you buy
- Count tasks, not people. Headcount does not price this product. Estimate how many research reports and automations a month your team will actually run, then multiply by 500 to 2,000 credits.
- Start on the free $100. It never expires, which makes it a real trial rather than a countdown.
- Watch the first full month, then pick a rung. Since rollover is capped at one month, buying ahead of demand wastes credits, and moving up a rung is a dropdown away.
- Ask about volume pricing at the point it matters. The discount sits behind Enterprise, so ask once you can see a month of your own burn rate.
How this compares to seat pricing
Credit pricing and seat pricing fail in opposite directions, and which one suits you is a question about your usage, not about which model is better.
A workspace price rewards a large team doing light work. Ten people asking Viktor a handful of questions a day is the case it’s built for, and no per-seat tool competes on that. A credit price punishes concentrated heavy use: a small team running real projects can pass what a per-seat tool would have charged, and the reported bills above are all from that end.
If your work is a steady flow of messages that need triaging, routing and turning into tracked tasks for named people, that’s a per-seat shape rather than a per-task one, and it’s what this+that is built for. If your work is a variable number of research jobs a month, credits will likely cost you less. Price the two against your own month rather than against each other’s headline.