Sage Oracle Jet Review – All-in-One Espresso Machine Tested

The Sage Oracle Jet espresso machine goes halfway towards a bean-to-cup machine. The Oracle Jet takes care of tamping as well as grinding. And on-screen prompts guide you through the workflow.

Our team of experienced baristas, trainers and café owners put the Oracle Jet through its paces for several months with two test units. The question: can this semi-automatic espresso machine with built-in grinder and modern features really make the switch from a bean-to-cup machine easier while delivering barista-level espresso?

At first glance, the spec sheet sounds promising: extremely fast heat-up, automatic grinding and tamping, a touchscreen with drinks menus (including cold brew) and an automatic steam wand for microfoam. All of that packed into one compact machine – but does it really work as well in practice as it does on paper? In this review we take a sober look, with our usual friendly wink.

Look forward to clear chapters on exterior and build quality, heat-up time & energy consumption, temperature behaviour, grinder, usability, espresso quality, milk frothing and the cold brew function – and at the end, our verdict with ratings.

Exterior and build quality

On the outside, the Oracle Jet is obviously a Sage. Stainless steel look, cleanly designed housing, large water tank at the back – classic Sage. Still, it’s clear that the Oracle Jet sits a step above the cheaper models (e.g. Barista Express/Pro), much like the Sage Dualboiler. We noticed higher-quality build: less plastic, more metal, no sharp edges – overall, everything feels solidly made. The machine weighs around 12 kg and stands firmly as a result. A handy detail: there are small wheels under the drip tray, so you can easily pull the 12 kg forward on the worktop, for example to top up the water or clean behind it. Little things like this show that Sage has thought about everyday use.

The dimensions are considerable – roughly 42 cm wide, 38 cm deep and 36 cm high – thanks to the built-in grinder and extensive features. The Oracle Jet is essentially a small bean-to-cup machine in portafilter form, and it needs some space. On the plus side: the water tank holds 2.3 litres and can either be opened from the top (handy if there’s little clearance under wall cupboards) or removed completely towards the back. The drip tray is generously sized and has a clever float that shows “tray full”. It holds about 650 ml of liquid before you should empty it – enough that you don’t have to run to the sink after every shot. Thanks to a plastic support plate under the stainless steel grate, not much sloshes over when you carry it. And at the front, behind the drip tray, there’s a drawer for accessories (blind basket, brush etc.), as we know it from Sage. In short: build and design are well thought out and made for daily use.

A big plus in this class: 58 mm portafilter diameter. Sage fits a standard stainless steel portafilter with a double spout and a large double basket. Unlike smaller Sage models (which often have 54 mm), we could use our usual 58 mm tampers, baskets and accessories. The double basket is fairly roomy – ~22 g of ground coffee fits in, and that’s what the machine suggests you use. For the best repeatability, Sage recommends (and we agree) brewing mainly double espressos.

Overall, the Oracle Jet makes a quality impression. We rate the material quality at 6.5/10 points – solid, if not fully at pro level. The package is complete: besides the machine with its integrated tamper, you save yourself buying a few extras. Included are, among other things, a knock box (for the spent coffee pucks – very commendable!), a 58 mm portafilter with double basket, a milk jug, water hardness test strips, a water filter for the tank and various cleaning tools (cleaning tablets, small brushes, a little screwdriver).

Heat-up time and energy efficiency

Now for one of the Oracle Jet’s highlights: its lightning-fast heat-up time. The first time we switched it on, we could hardly believe it – after 3 minutes the machine was ready to go! That’s close to a record in the world of espresso machines. For comparison: many classic heat exchanger or dual boiler machines need 20–30 minutes, and some thermoblock machines still take 5–10 minutes. The Oracle Jet sets a new benchmark here, and with its 3-minute heat-up time it ranks among the top 3 machines we’ve tested so far. Sage calls the heating system ThermoJet® – it consists of two digitally controlled thick-film heaters. One quickly heats the water for espresso or steam, the second is built directly into the group head to bring the brew group up to temperature. This design explains the speed as well as the good energy efficiency: the heaters have little mass, which is heated in a targeted way, instead of keeping a large boiler hot all the time.

Our measurements impressively confirm the efficiency. In the standardised test (heat-up + 5 espresso shots in a row), the Oracle Jet used only about 0.0656 kWh – in other words, under 0.07 kWh. Broken down to a single espresso (including the heat-up phase), that’s a tiny 0.014 kWh per shot. Put differently: one espresso uses less electricity than, say, 5 minutes of ironing with a steam iron.

With energy prices rising, that’s reassuring news: your daily espresso habit on this machine will barely register on your electricity bill. With 10/10 points in the power consumption category, the Oracle Jet sets a benchmark. Once heated up, the machine doesn’t keep reheating either – it switches to standby fairly quickly when idle and apparently holds its temperature efficiently without drawing power constantly.

For you as a user, all of this means: no more long waits in the morning – switch on, fill the beans, maybe set out the cups in the meantime, and off you go. Sage somewhat misleadingly advertises “3 Second Heat Up” on other models – in reality it’s 3 minutes, but that’s still sensationally fast.

Still, here’s a tip: as quickly as the Oracle Jet gets hot, 3 minutes aren’t enough for perfect temperature stability (details in the next chapter). After the 3 minutes, ideally run a flush or at least rinse hot water through the portafilter briefly. That heats the portafilter through properly. The metal parts (portafilter, brew group) lag well behind – that’s the case with every fast heat-up machine.

All in all, heat-up time and efficiency are the Oracle Jet’s absolute strengths. That “bam – machine on and straight into it” feeling is genuinely fun and lowers the bar for pulling a spontaneous espresso. For this we award full marks, 10/10, for heat-up time. But now for the flip side of this speedy ThermoJet technology: temperature behaviour during the espresso shot.

Temperature behaviour during the shot

As quickly as the Oracle Jet is ready on paper, we have to take a critical look at its temperature stability during extraction. Our tests showed that the machine does reach the set temperature, but only with a delay into the shot. What does that mean in concrete terms?

The Oracle Jet does have an adjustable brew temperature (configurable in °C on the display, around 93 °C out of the box). Our measurements showed no big offset between the displayed target temperature and the actual temperature – that’s a positive. If you set 93 °C, for example, the effective water temperature (after the initial phase) settles at roughly that level. But: for the first few seconds of every shot, the water runs through noticeably cooler.

There are two reasons for this. First, the machine doesn’t start extracting immediately after you press the button. Sometimes a few seconds pass before the first drops come out of the machine.

The second point comes down to the machine’s extraction logic. The Oracle Jet has a pre-infusion stage with very low flow. It uses a vibration pump, but it is electronically controlled so that it only builds full pressure after about 8–10 seconds. At first, only a thin trickle of water drips onto the coffee bed – which can in principle be good for extraction (gentle wetting of the grounds). The problem: this small amount of water cools down a lot in the measuring portafilter during pre-infusion, and the heating system apparently can’t bring the water to target temperature from the very first second. Result: the first ~10–15 seconds of the shot run below target temperature; only from around second 15 does the water flow reach the target.

This too-low temperature is also partly down to the design of the measuring portafilter. At such a low flow rate, it doesn’t come into contact with enough water to measure high water temperatures. On the other hand, such a small amount of water also cools down extremely quickly. We recreated this low flow on other espresso machines. The direct comparison shows: the Sage stays far too cold for far too long in every case. All the other machines with low flow reached the target temperature 5–7 seconds faster, at the same flow rate!

We recorded this behaviour with both our KM temperature protocol and the WBC standard protocol (32 s shot with 5 s pre-infusion). The range was huge: we measured water temperatures between 66 °C (minimum) and 94 °C (maximum). Even ignoring the first 5 seconds, the temperature at the start was still in the 80s and only climbed above 90 °C late on. This “sluggish temperature curve” also shows up in our own protocol. After 4–5 seconds of the shot, the temperature was still around 5 °C below target, and after 10 seconds still about 1.5–2 °C under. And all that on the “flattering” curve, for which we only started counting the shot once 70 °C had been reached.

When we measured from the moment the button was pressed, which is what the test protocol normally specifies, we only reached the target temperature after around 18–22 seconds!!!

In both cases, the temperature only becomes stable after that. That’s far from ideal, because the decisive extraction phase for espresso happens in the first 20–30 seconds. It doesn’t help much if the machine is rock-steady from second 20 onwards.

For us as testers, this is one of the biggest points of criticism of the Oracle Jet. Sage has evidently taken a defensive approach here: better a lower temperature at the start than risky overheating. However, that leads to inconsistent extraction – especially with lighter roasts, which need high and constant temperatures. Our temperature curves also showed slight fluctuations from shot to shot: some shots ran relatively constant (after the initial rise), others varied by 1–2 °C. Overall, the temperature profile remains “problematic” and far from ideal. We awarded only 4.5/15 points in the temperature stability category – a clearly below-average score that drags down the overall rating.

To put this into perspective: other machines do better, either reaching temperature faster or rising slightly over the shot, but from a higher starting level. And we’re not talking about machines in the same price range, but about espresso machines that cost a third of the Sage Oracle Jet.

The Oracle Jet, by contrast, runs a “too cold for too long, then constantly a bit higher” profile. That’s neither a flatline temperature nor a classic temperature profile we would want. Our view: there’s clear room for improvement here. Thick-film heater technology can easily handle it.

Practical temperature tips

Let it heat up and flush: As mentioned in the previous section, be sure to wait the 3 minutes and run one hot flush to preheat the portafilter and brew group as much as possible. A “thermal flush” before the first shot gives you a better thermal start.

Preheat the cup: Of course, your espresso cup should be well preheated (unfortunately the Oracle Jet’s cup tray doesn’t get hot, as there’s no boiler, so rinse it with hot water or warm it on a separate hot plate). Every bit of stored heat helps.

Set a higher brew temperature? You might think you could simply set it 2–3 °C higher to compensate. We tried that – but it causes the later seconds of the shot to overheat. So it’s better to stick with a sensible setting (e.g. 93 °C) and adjust the coffee blend instead.

Choice of beans: We recommend using medium to darker roasts with the Oracle Jet. They are more forgiving of the slightly lower starting temperature and still extract harmoniously. In fact, we got excellent results with an Italian espresso blend we developed ourselves (darker, full-bodied, chocolatey). Light, acidity-driven third-wave roasts were harder to get properly sweet and clean. With a strong, full-bodied espresso, on the other hand, the result tastes good, because this coffee masks the temperature weakness somewhat.

Don’t stretch the shot time unnecessarily: You might be tempted to choose very long shot times so that the temperature eventually gets there. We advise against it – an espresso shouldn’t run for 50 seconds just to hit 93 °C at the end. Stay in the normal range (25–35 seconds total shot time). A far-too-long shot brings nothing in the cup – it just leads to over-extraction and bitterness.

In summary: temperature behaviour is the Oracle Jet’s biggest weakness. For absolute espresso perfectionists expecting a flat 93 °C curve, it will be disappointing. In practice, though, using the measures above, we could still pull tasty espressos – especially if you meet the machine halfway with suitable beans. From halfway through the shot, the set temperature roughly matches reality and has no big hidden offset, which is a positive. Still, it’s a clear deduction for style. In our rating criteria, temperature was the low outlier with just 4.5 out of 15 points.

Built-in grinder – performance, noise and grind quality

A central part of the Oracle Jet is its built-in coffee grinder. In many compact machines, the grinder is the sticking point: often a compromise for reasons of space. So we applied our full grinder test protocol to check noise, grinding speed, dead space and particle size distribution – just as we do with standalone grinders.

The Oracle Jet uses hardened steel conical burrs. Sage says they are Etzinger burrs (known from Baratza grinders), which our visual inspection confirmed. The conical burrs are adjusted steplessly, but with numbered settings 1–45 shown on the display (so the fineness can be set in 45 “steps”, with some leeway between the numbers – the system is effectively “micro-stepped”, which allows fine adjustment). In use, the grinder automatically grinds straight into the portafilter and then tamps automatically as well. But let’s look at the numbers:

Grinding speed: The grinder is neither a snail nor a turbo, but sits comfortably mid-field. We measured about 18 g in 10 seconds. That’s perfectly fine. You won’t be waiting forever, but it’s not all done in 5 seconds either. For home use, the speed is more than enough.

Noise: At an average of 75.9 dB while grinding (measured from 30 cm away), the Oracle Jet’s grinder is pleasantly quiet. Our subjective impression confirms this: the grinding noise is more muffled than with many external grinders. The machine itself (pump, mechanics) is fairly quiet anyway – the most noise always comes from the grinder, and here the Oracle Jet performs very well. 75–76 dB puts it in a very good category (for comparison: that’s well below the noise level of many cheaper grinders, which easily reach 80–85 dB). Nobody will jump out of bed in shock when you grind an espresso at home in the morning. One more word on the “quality of the noise”, though: the grinder isn’t loud, but while grinding it sounds like it’s “straining” to get the coffee fine.

Dead space (“retention”): Less pleasing is the dead space, i.e. the amount of ground coffee that stays in the grinder after grinding. We normally distinguish between permanent and temporary dead space – but with this design it couldn’t be broken down cleanly, so here’s just the absolute figure: about 3.3 g stays stuck in the grind chute and between the burrs. In practice this means: if you don’t grind any coffee for a while (say >1 hour), the first 3.3 g that come out afterwards are already slightly oxidised, stale grounds from last time. You won’t necessarily taste that dramatically straight away (especially not in milk drinks), but over time you’ll notice that the first shot after a long break tends to be worse and runs a bit thinner than the second. If you care about freshness, that’s not ideal. Unfortunately, you can’t magic away 3 g of old coffee either – unless you do a short “purge grind” and throw away the first few seconds of grounds. Hardly anyone will do that day to day, so you live with slight quality differences between the first and subsequent shots. For a built-in grinder with doser and tamper, though, 3.3 g isn’t entirely surprising – classic flat burr grinders with dosers used to have much higher retention. Modern single-dose grinders, on the other hand, are under 1 g. Let’s be clear: by design, a built-in grinder can hardly be made as low-retention as, say, external single-dosing grinders, where you drop the beans in at the top and everything comes out at the bottom.

Grind consistency & particle size distribution: Now it gets technical: we had the particle size distribution of the grounds measured by laser diffraction – as with more than 80 grinders before. But don’t worry, we’ll just summarise the relevant results. At an espresso setting, the Oracle Jet grinder produces a median particle diameter of about 324 µm (x50). More important is the width of the distribution: how uniform or varied are the particle sizes? The main peak (the tall “mountain” in the particle size chart) was at ~284 µm with comparatively narrow spread. The share of very fine particles (<100 µm) is 28.5 %. In plain terms: the grounds are relatively uniform, there aren’t excessive fines, and there’s no extremely wide spread from very fine to very coarse.

Our grinder tests so far have shown that a narrow particle distribution contributes significantly to a clear, balanced espresso taste. Widely spread particles (very uneven grounds) often lead to “frayed” flavours – bitter notes AND unpleasant sourness at the same time, little sweetness, simply unbalanced. The Oracle Jet grinder holds up pleasingly well here.

The Oracle Jet’s particle curve resembles that of good mid-range espresso grinders. It isn’t as narrow as with really expensive high-end grinders (there the peak would be even narrower and more defined), but “good and solid” sums it up. In our trials, many different espresso roasts could be extracted tastily without the grinder standing out as the limiting factor. We even managed to extract lighter roasts – only, as mentioned, it’s the temperature rather than the grinder that limits things here.

We give the grinder good marks for noise, speed and consistency (e.g. noise 3.5/5 points in our scheme, speed “solid”). The dead space pulls it down a little. Overall, the grinder scores “good” – which you can’t say of every built-in grinder. We’re normally not fans of combo solutions (grinder + machine in one), because at least one component often suffers. Here, though, the grinder isn’t a big weakness but rather a small strength of the Oracle Jet.

Tamper

A word on the Oracle Jet’s automatic tamper: after grinding, the machine automatically compresses the grounds. You take out the portafilter and find a ready-tamped puck surface. At first we were sceptical – we’ve seen worse versions of this kind of solution. But Sage has done it well: the pressure is consistent, the puck neatly compacted. Just a few loose crumbs sometimes remain at the edge, but you know that from tamping by hand too (e.g. if the tamper is slightly smaller than the basket diameter). You can ignore these crumbs – they don’t noticeably affect extraction. Important: the tamping pressure can’t be adjusted, but it is apparently preset in the optimal range. All in all, the auto-tamper saves work and works reliably.

In short: automatic tamping is a nice thing, especially if several people use the machine or you want to work precisely in the morning without much effort.

Usability and user guidance

The Sage Oracle Jet presents itself as a “smart” espresso machine – and you notice that in how it operates. The central element is the large 5-inch touch display in the middle. We really liked this display: it’s clear, in colour and above all fast and responsive. Nothing is more frustrating than a sluggish touchscreen – here everything responds well to touch.

User interface and menus

After switching on, the Oracle Jet greets you with a home screen showing various drink options as icons. By default, espresso, americano, cappuccino, latte, flat white etc. are listed – essentially predefined programmes. But you can personalise the order: move your favourite drink to the front, hide rarely used ones (they then disappear from the first view but aren’t deleted). Dragging and dropping the icons felt intuitive in our test.

Tap a drink (e.g. espresso) and you get to a submenu with specific settings. Here you can set, for example, the volume (i.e. how many ml/grams should go into the cup), plus, if needed, the brew temperature and pre-infusion time (depending on the mode). For milk drinks, options for milk foam amount and milk temperature are added. For a cappuccino, for example, you can define: X ml espresso, foam level 5, milk temp 60 °C. You can save these adjustments as a personal preference so the machine uses these parameters straight away next time. You can even change the drink’s icon – a gimmick, but a nice one.

In the menus, Sage has tried to strike a balance: as many settings as possible, but still simple navigation. In our view, it has succeeded. Sure, at the start you might need a look at the manual to understand all the features – but we found our way around quickly. If you’re coming from a bean-to-cup machine, you’ll be pleasantly surprised by how much you can adjust. If you’re coming from manual machines, you might think “is all this really necessary?” – but don’t worry: you can also simply operate the Oracle Jet like a classic machine (adjust grind size and dose manually, then start the shot) without constantly bothering with menus. For shared households or beginners, though, the helpers are worth their weight in gold.

Barista Guide and assistance

One special feature is the built-in “Barista Guide”. It’s basically a little helper that can give feedback after an espresso shot: was the shot under- or over-extracted? The machine detects this from the flow time and volume and then gives a tip on which way to change the grind size. Example: if your espresso ran through far too fast, after the shot you’ll see the recommendation “grind finer, e.g. from 30 to 27”. Or, the other way round, grind coarser after a very long shot. We found this feature useful – especially for people who don’t have years of espresso experience yet. The Oracle Jet takes you by the hand and may save you the odd frustrating shot. Of course, it doesn’t replace a barista course, but it helps you learn faster: you get a concrete suggestion for what to do, instead of just noticing “damn, ran too fast/slow”.

The Oracle Jet has precise volumetrics: a flow meter ensures that, for example, 45 ml of water runs through and then stops. We tested the consistency – the machine hits the volume quite accurately (within 1–2 ml). That’s why it gets 8/10 points in the “volumetrics (consistency of output volume)” category, a good score. We advise against time-based stopping, on the other hand: if you set a fixed 25 seconds, depending on the grind size the cup might overflow one time and stay almost empty the next. The result would be inconsistent. So definitely work volumetrically! After all, the volumes can be set in the profile. Unfortunately, though, the control can’t be customised here. It follows the volumes specified by Sage, which do make sense in relation to the large dose of coffee used.

We rated overall usability at 8.5/10 points. Despite its many features, the machine is intuitive. Much of what sounds complex runs automatically or is optional. Example: the auto-queue function – the Oracle Jet can optimise the workflow by, for example, automatically starting to froth milk as soon as the espresso has finished extracting, without you having to press anything. That way the espresso doesn’t sit around for too long; the milk is ready right after, and you can combine the two. It’s well thought out and saves a few seconds – nice to have when you’re making cappuccino & co.

Cleaning and maintenance are well supported too: the machine reminds you about descaling, brew group cleaning (backflush) and changing the water filter, each shown in the menu. There are automated cleaning programmes you can start. All of this means that less experienced users can also keep the machine in good shape for a long time. Still, an important note: be sure to use the supplied water filter (or fill it with soft water)!

Another issue that stood out in use: the portafilter’s double spout. Here we have a clear point of criticism: the two spouts are very close together at the bottom. Although they start far apart at the top of the portafilter, they taper inwards so that the two espresso streams are only a few millimetres apart as they come out. Why does that matter? If you want to fill two espresso cups at the same time, they have to stand really close together, otherwise some goes to waste. A few times we were annoyed that a drop landed next to the cup when splitting a double shot, because the spout is so tightly centred. It’s not the end of the world – you get used to it and position the cups accordingly – but we noticed it negatively. On many other portafilters, the spouts are a bit further apart at the bottom, which makes it easier to pull two espressos. Alternatively, you could use a bottomless portafilter (available as an accessory); then you don’t have this problem, but you have to split the shots manually.

All in all, the Oracle Jet convinces on usability. The mix of automation (for convenience) and manual control (for quality) is well judged here. We’d say: for someone coming from a bean-to-cup machine, a lot feels familiar (“press a button, it runs automatically, you get hints”), and at the same time you gradually get to know the important parameters of an espresso machine (grind size, ratio, temperature etc.). Advanced home baristas can enjoy it too, because despite all the helpers you can adjust enough yourself, and the machine never “gets in the way” when you want to step in yourself. For example, you can grind and tamp manually at any time (the grinder also has a manual button, and the steam wand can be operated completely manually, see next section). In short: usability = a success.

Espresso quality in the taste test

In the end, what counts is what ends up in the cup. How does the Oracle Jet do on taste? Naturally, all the aspects so far come together here: temperature, grinder, volumetrics – everything affects the espresso. Our test results were good overall, with certain limitations.

With the right bean (as mentioned, not an ultra-light third-wave light roast, but something with a medium to darker profile), we were able to pull lovely espressos. Crema, body and aroma were at the level we expect from a good espresso machine. Especially with our darker Compadre espresso (an Italian-style roast we developed ourselves), we got a thick, hazelnut-brown crema and a full body with notes of cocoa and nuts, and the temperature issue wasn’t noticeable in the cup – the espresso was sweet and strong, just as it should be.

With lighter espressos (fruit-forward beans), however, we hit clear limits. Here we noticed that the machine doesn’t quite achieve the optimal extraction that would be possible with a precise, temperature-stable espresso machine. The shots varied a lot, from drinkable to extremely sour, and sometimes lacked clarity, body and balance. We attribute this primarily to the temperature curve – slight under-extraction in the first phase, which you perceive in the cup as too much acidity or a thin taste.

Espresso potential is what we call one of our rating categories – i.e. the maximum quality you can get out of the machine. Here we give the Oracle Jet 6/15 points. It’s above all the temperature performance that severely limits the machine. We also miss precise settings for volume (water) and coffee (less than 20 grams, fiddly adjustment with the propeller). For the home barista who likes to experiment, that could be frustrating – but then again, that’s exactly the target group a partly automated machine isn’t primarily aimed at.

In terms of repeatability, the Oracle Jet did well. Thanks to the volumetric control and consistent grinder, consecutive shots (with the same settings) were very similar. What matters, of course, is heating the machine up properly and bearing in mind that the water sometimes starts flowing with a delay of 2–4 seconds.

When making several espressos in a row, the quality was consistent. This is where the electronics pay off: every double shot came out at pretty much ~45 g in ~25–30 seconds, as long as the grind size was right.

Milk frothing: automatic vs. manual

For fans of cappuccino, latte macchiato & co, the Oracle Jet has another ace up its sleeve: an automatic steam wand with a milk foam programme. Sage calls the system Auto MilQ™, and it promises barista-level milk foam at the touch of a button – at least in theory. We tested it extensively, of course, with both cow’s milk and oat drink, and always compared it with milk foam we frothed by hand.

Automatic milk foam

The Oracle Jet has a stainless steel steam wand fitted with sensors. You put the wand into the jug (filled with milk), start the automatic programme on the display and can then watch the machine froth the milk. Beforehand, you set two things: the milk’s target temperature (selectable between 40 °C and 75 °C) and the foam texture (level 1–8). Behind the scenes, Sage has also stored profiles for different types of milk – for example, there are specific settings for soy milk, almond milk etc. that you can select so the automatic mode froths to match. That’s pretty well thought out.

In practice, automatic frothing went like this: first the wand hisses air into the milk (baristas call the phase in which coarser bubbles are created the “stretching phase”). Then, after a while, the air injection stops and the milk is only heated and swirled (“texturing phase”). The Oracle Jet does all of this in a fixed sequence. Our impression: the result was “okay to good”, but not perfect. Why? We noticed that the machine keeps the texturing phase relatively short, and starts it late.

When frothing milk, we like to aim for temperatures just under 60 °C. That’s when milk is very sweet and, in our view, tastes best. But the Auto MilQ’s foam structure is set up so that it still needs temperatures just above 60 °C to roll the milk really evenly. So if you want to froth your milk automatically, you have to accept a small compromise here.

We also noticed that the automatic mode didn’t produce exactly the same amount of foam every time. Even though we always filled the jug to the same mark, there were slight differences in foam height – nothing dramatic, but noticeable. It may be down to minimally different starting temperatures of the milk or the nozzle setting. Either way, the consistency of the auto foam was “okay, with small fluctuations”.

Manual frothing

Fortunately, the Oracle Jet also lets you take over manually. You can switch the steam wand to manual mode by tilting it forward – the steam function then starts and you control it yourself like on any classic machine (by opening/closing a valve, or in this case by moving it back). Of course, we also frothed manually to check the steam wand’s performance. And lo and behold: the power is absolutely sufficient to bring 200 ml of milk up to temperature quickly, and with a little practice you can create excellent microfoam. The steam power is roughly on the level of good single boilers or thermoblocks. Important: the Oracle Jet doesn’t have a dual-heater system, i.e. for steaming, the ThermoJet element is heated to steam mode. But that takes barely a few seconds. After the espresso shot, you can steam practically straight away (the machine shows when it’s ready). Conversely, after steaming it has to cool down again for the espresso shot – this happens automatically by flushing into the drip tray. Overall, the switchover time is very short, and we didn’t find it bothersome.

Our comparison of automatic vs. manual milk foam:

Automatic: delivers consistently good but rarely perfect foam. Someone without practice will probably make a better cappuccino with it than with a simple entry-level espresso machine where you have to froth yourself. It’s practically beginner-proof: no scorched milk, no guesswork. You press start and the result is usable every time. That’s a clear advantage over bean-to-cup machines, whose foam is often much coarser and cooler – here you already get something close to a cappuccino made by a barista. We’d say the Oracle Jet’s auto foam is better than 90 % of what bean-to-cup machines produce, but still some way off a top barista result.

Manual: lets experienced users get the very best out of it. We managed nicer foam manually than the automatic mode, simply because we could hold the texturing phase longer and dose the air more precisely. With the right jug position (tip: wand close to the edge, almost touching the bottom, and let the milk swirl), we achieved fine, glossy microfoam, perfect for latte art. However – and to be fair, this has to be said – that takes practice. A beginner won’t get perfect foam straight away without guidance. This is where the advantage of the automatic mode comes back into play: it gives everyone at least drinkable milk foam, no training needed.

Also handy: the steam wand cleans itself automatically. After every frothing cycle, it briefly blows out steam and water to flush milk residue from the tip. You give the outside a quick wipe with the supplied damp cloth – done. The machine shows the maintenance interval for a more thorough clean (rinsing with cleaner).

Cold brew & cold drinks

New on the Oracle Jet are special programmes for cold coffee drinks. Sage has noticed that cold brew and cold coffee are trending and has therefore given the Oracle Jet a “Cold Extraction Enabled” stamp. This means: you can select “Cold Brew” or “Cold Espresso” in the menu, and the machine then tries to extract coffee with cold water.

What happens in cold brew mode? The machine grinds relatively fine (it even asks you to adjust the grind size accordingly; a prompt pops up saying “Please grind finer for cold extraction”). Then it starts a very long shot at low pressure. The aim is to make a kind of concentrate that you can then pour over ice. To be honest, we were sceptical – cold brew is normally made with a coarse grind steeped in cold water for many hours. A 5-minute extraction with espresso grounds? Sounds wrong.

The result was surprising: it was drinkable and not as bad as expected. We tried a cold brew shot served on ice and stretched with a little water – it actually had less acidity and bitterness than a hot-brewed espresso over ice (which often has “shock brew” effects).

To sum up: We genuinely enjoyed the Sage Oracle Jet in our test. Despite our critical analysis of the measurements, we enjoyed many espresso and cappuccino moments with it day to day – and that’s what counts. It’s a convenient machine that doesn’t keep you waiting and does a lot for you, while still producing real espresso that can taste really good with a bit of fine-tuning.

For us, the target group fits: if you’re coming from a bean-to-cup machine and want real espresso, but still value automation and convenience, the Sage Oracle Jet could be the right machine for you.

Have you already had experience with the Sage Oracle Jet? Are you perhaps exactly the target group looking to switch from a bean-to-cup machine? Let us know and share your opinion – praise or criticism – in the comments. We’re looking forward to hearing what you think!

With that in mind: happy brewing – hot or cold – and see you at the next test!