Bellwether vs. Probat: Which Roaster Fits Your Space?

You are pricing a roaster and Probat is on the list, which is exactly where it should be. It is the most established name in the category and by its own account seven of every ten cups in the world are roasted on Probat equipment. Nobody gets fired for buying one.

The useful question is not which machine is better in the abstract. It is which one fits the room you actually have, the staff you actually employ, and the volume you actually roast. On those three, the two machines separate very quickly.

The honest starting point: these are not competitors

A Probat is a gas drum roaster built for a roastery. Pricing runs from roughly $50,000 to $200,000+ depending on capacity, and that is the machine alone. It expects a gas line, an exhaust path, and somebody who knows how to drive it. In a production roastery with a roasting team, that is all completely reasonable, and the machine rewards the skill you put into it.

The Bellwether Shop Roaster is $14,900, or $19,900 with continuous roasting. It is electric, it is ventless, and it runs a profile automatically. It plugs into the same kind of 240V outlet as your espresso machine. Where a Probat asks for a building, this asks for a countertop and a circuit.

So the comparison is not really Probat versus Bellwether. It is: do you want to build a roastery, or do you want to roast in the space you already lease?

What happened when Square Mile put them side by side

This is the part worth dwelling on, because it removes the question most people actually want answered — whether an automated electric machine can hold its own against a production drum roaster.

Square Mile Coffee Roasters in London is one of the most respected specialty roasters in the world. They tested a Bellwether alongside their production Probats and cupped the results blind with their own team.

“We put our Bellwether roast on as well as production roast from our Probat machines. About 20-25 of our team were tasting. And no one could pick the production roast from the Bellwether roast. Most of them were the Bellwether roast as their favourite. That was a nice win.”

Tom Flawith, Head of Coffee, Square Mile Coffee Roasters

That is a room full of professional palates, tasting blind, against their own daily production standard. It does not make a Bellwether a Probat. It does mean the quality question is settled well enough that you can decide on the practical grounds instead.

What your roaster actually does

There is a version of this comparison where employing a skilled roaster settles it in favour of the drum machine. It does not, and Square Mile is the reason it does not — they are a roastery full of roasters, and what they described was not deskilling. It was the work moving up a level.

“The operating of it was very straightforward and very easy to use. That allowed us to focus on the sensory side of things. The machine kind of took care of the roasting part. We could then just focus on making the coffee taste as good as it could be.”

Tom Flawith, Head of Coffee, Square Mile Coffee Roasters

The curve still has to be designed by someone who knows what they are doing — profile it, cup it, correct it, cup it again. What changes is what happens after that. On a drum roaster your roaster executes that curve, batch after batch, and the result carries their attention that day. On a fleet, they design it once and every machine runs it identically — which is worth more, not less, the more machines you have.

Recent Coffee Roasters went in with exactly this worry and came out the other side of it:

“I was genuinely nervous about losing the artistry. And actually we gained. Because we were able to control the coffee. The thing that scared me the most was a customer coming in for the first time and getting a bad batch. And now that almost can’t happen.”

Barry, Recent Coffee Roasters

It is also worth saying plainly that automation is not the dividing line here. Probat, Loring, Diedrich and Giesen have all shipped automation on their newer machines. The difference is not whether the roast is programmed — it is whether the building needs gas and a stack for it to run.

Where a Probat is the right machine

• You want the whole volume in one machine and one operator, rather than several running side by side.
• You already have gas, an exhaust path, and a permit history in the building.
• You want to drive each batch by hand in the moment — reading it and adjusting by feel as it runs — rather than designing a profile and having it executed the same way every time.
• You are building a dedicated production facility rather than roasting inside a retail space.

If three of those four describe you, buy the Probat. The rest of this page is not for you.

Where the Bellwether is

• You are roasting in a café, a restaurant, a bakery, a hotel, an office building or a historic property where an exhaust stack is not happening.
• You do not have a roaster on payroll and do not want to hire one — standard roasts need about two minutes of attention, and staff learn it in under half an hour.
• You want the machine in front of customers rather than in a back room.
• You are replacing wholesale rather than scaling a production line.

Scaling by adding machines, not by buying a bigger one

The assumption behind most roaster comparisons is that growth means a larger drum. On a gas roaster it has to — every size increase is a new machine, a new install, and usually a new room. That is the model a Probat is built for, and it is why capacity per machine is the number everyone compares.

A ventless electric machine changes the shape of that decision, because the second unit costs the same to install as the first. There is no gas line to extend, no second hood, no additional rooftop penetration and no new permit path — each one needs a 30-amp circuit. Operators who grow past a single machine tend to add units rather than replace them, and several of ours run banks of them side by side.

That has consequences a single larger machine cannot match:

• Different coffees roast at the same time on different profiles, instead of queueing behind one drum.
• A machine down is not production down. On a single-roaster setup, a service visit stops everything.
• Capacity arrives in $14,900 steps rather than one large capital decision, and follows revenue instead of anticipating it.

It also goes further per machine than most people assume before they run one. Anchor and Tree runs a wholesale business on Bellwether and put it plainly:

“I am doing between 3,000 and 4,000 pounds a month as a wholesale coffee roaster, and I still have extra time to roast.”

Donovan, Anchor and Tree Coffee

And for multi-site operators the distributed model is the point rather than a compromise. Recent Coffee Roasters built their whole business on it:

“I fell in love with the idea of how the Bellwether fit our model of decentralized multi-site roasting and how we could program everything in one space using one team, knowing that we were getting the same coffee throughout all the sites.”

Gaz, Recent Coffee Roasters

So volume alone does not decide this. What decides it is whether you want that volume consolidated into one machine in one room, or distributed across several that can go wherever you have a plug.

The roastery stops being a place

This is the part that has no equivalent on a drum machine, and for a multi-site operator it is usually the whole argument.

A central roastery is a hub with a distribution network attached. One building roasts, and every bag then travels to wherever it is going to be drunk. That network is a permanent operating cost, and it runs against a clock — roasted coffee is at its best for a couple of weeks, so deliveries have to be frequent, fast and on a schedule you cannot let slip. The further a site is from the roastery, the worse the coffee is when it arrives, and the more it costs to get there.

Distributed roasting inverts that. Your roaster still designs every curve — profiled, cupped, corrected, signed off — and then those profiles run on machines standing in each of your locations. What moves between sites is green coffee, and green has no clock on it. It keeps for months, so it can travel slowly, in bulk, infrequently, by the cheapest freight you can find. Roasted coffee is the thing that has to move fast and often. Stop moving it and that whole cost structure goes away.

What each site gets instead is coffee roasted where it is served, on the day it is served. "Locally roasted" stops being a claim about your head office and becomes literally true at every address.

“Wholesale customers love knowing their coffee is roasted just down the road.”

Gaz, Recent Coffee Roasters

Consistency is the thing people assume you give up here, and it is the thing you keep. The profile is the same file on every machine, so the cup does not drift by location the way it does when different sites are supplied by different roasters — or by the same roaster on a different day.

“If you have one in each shop, from wherever you’re at, you have your master roast. Boom, easy. And every, you know, you have consistency across the nation.”

Elizabeth, Pitchside Coffee

And because a site only needs single-phase power, the map opens up much further than a delivery radius. High Grade have thought this through further than most — the constraint they are describing is not distance, it is customs:

“We’ve seen how this could be repeatable or replicated in either multiple sites, remote wholesale applications, opportunities for cross-border selling without having to cross the borders. We could set up high-grade entities in Europe where the roasting all happens there, product closer to its final customer.”

Liam, High Grade Coffee

Which is the modular version of growth. A new location is a machine and a circuit, not a larger central roastery, a bigger warehouse and another van on the road. Capacity arrives with the footprint instead of ahead of it, and nothing about site eleven is harder than site two.

The install question nobody prices upfront

The sticker price is the part everyone compares. The install is the part that decides the timeline, and on a gas roaster it is frequently larger than people expect:

• Gas line installation: $5,000 to $15,000
• Exhaust hood and ductwork: $3,000 to $10,000
• Afterburner or thermal oxidizer: $10,000 to $30,000
• Rooftop penetration: required, and often the item your landlord refuses
• Timeline: four to twelve weeks, plus permits

The Bellwether needs none of it. Smoke and VOCs are neutralised in an internal afterburner, so the air leaving the machine is already filtered and nothing vents outside. What it needs is a 200 to 240V, 30-amp single-phase circuit — an electrician typically installs one in two to four hours for $500 to $2,000. Then you plug it in and roast the same day.

That gap is also why the two machines end up in different buildings. Ventless is not a specification so much as a permission slip: it unlocks locations — mall food courts, office cafés, historic buildings, shared kitchens, airport terminals — where a vented roaster is simply not an option at any price.

The parts that are not about the machine

A Probat is hardware. What arrives with a Bellwether is a program: a green coffee marketplace where the coffees come with roast profiles already dialled in, a coffee team who help you choose your first coffees and build a menu, packaging support if you want to sell retail bags, and preventive maintenance at six and twelve months. If you already have a green buyer and a roaster on staff, that is worth little to you. If you do not, it is most of the job.

So which one

If you can vent, you are building a dedicated roastery, and you want it all in one drum — buy the Probat. If you are roasting inside a retail space, or you cannot run gas, or you would rather add capacity in steps, or you are serving more than one address and tired of shipping roasted coffee to reach them — then the comparison resolves on your building, your growth curve and your map, rather than on the coffee. Square Mile already settled the coffee. Most operators save $1,000 to $5,000 a month against wholesale and pay the machine back in about six months.

Run your own numbers before you decide either way — volume, current cost per pound, and what your landlord will actually permit.