Amber Electric Review: Eight Months With Solar and a 48kWh Battery
I joined Amber on 11 January 2026 and added a 48kWh battery on 11 April 2026. Every figure below came out of Amber's own API rather than a screenshot of the app, and covers 87 days from 22 June 2026 to 16 September 2026. It is not the review I expected to write.
- Avg import price
- 14.16c
- Avg export price
- 11.70c
- Import in the 4–8pm peak
- 4.2%
- Would I stay?
- Yes
The short version
Amber has been the right decision for my household and I have no plans to leave. But the reason it worked is not the reason I signed up, and that gap is the most useful thing I can tell you.
Amber does not sell cheap electricity. It sells access to the wholesale price for $25 a month. Every other retailer buys on that same market and resells to you at a marked-up fixed rate, absorbing the risk in exchange for the margin. Amber removes the markup and hands you the risk. Whether that is a good trade depends almost entirely on whether you own something that can move energy through time.
What actually changed. Across 87 measured days my grid import averaged 14.16c/kWh. On my previous plan every unit cost 25.45c flat. My exports earned 11.70c/kWh against a 4c feed-in tariff. That is the entire case for Amber, and it has nothing to do with price spikes.
I expected to make money catching evening price events. I did not catch a single one. What happened instead was quieter and, in hindsight, more durable: the battery removed my household from the expensive part of the day almost completely, and the wholesale price for the rest of the day turned out to be far below what a retailer had been charging me for the same electricity.
What I am running, and why that changes everything
Every Amber review is really a review of the reviewer's house, so here is mine. Two adults in a two-storey home in southern Sydney, on the Ausgrid network:
- 10.66kW SunPower array, installed March 2023.
- FoxESS H3-10.0-Smart hybrid inverter and a CQ6 battery stack of roughly 48kWh, installed 11 April 2026, replacing an earlier Sungrow inverter.
- Hot water on a controlled-load circuit — which turns out to matter enormously, for reasons I will come back to.
- A smart meter, which Amber requires.
That battery is large. Most Australian home batteries are 10–15kWh, and most Amber reviews are written from that position. My results are better than yours will be if your battery is smaller — a 48kWh battery can ride out an entire evening peak, soak up a full day of cheap midday power, and still have capacity spare. A 10kWh battery does one of those at a time. The model still works at that size; it just works less completely.
The timeline matters too. I was on Amber for three months before the battery went in, and those were the worst months I have had. January in particular was poor: no storage, summer prices, and SmartShift still learning my house. If you join Amber without a battery and judge it on the first quarter, you will conclude it was a mistake.
How Amber Electric actually works, in plain English
If you already understand the National Electricity Market, skip to the next section. If not, this is the part that makes the rest make sense.
Electricity in eastern Australia is traded on a live market. Generators offer power, the market operator works out how much is needed, and a price is set for your whole state, recalculated every five minutes and settled every thirty. That price moves enormously. On a sunny afternoon with more rooftop solar than anyone needs, it goes to zero and frequently below. On a still, hot evening when every air conditioner in Sydney is running, it can briefly reach thousands of dollars per megawatt hour.
Your normal retailer sees all of that and hides it from you. They buy on that market, hedge against the spikes, and sell you a flat rate that includes their margin and their insurance against a bad summer. The flat rate is a service, and you pay for it. In my case I was paying 25.45c/kWh for it, at every hour of the day, whether the underlying energy was worth two cents or eighty.
Amber does the opposite. You pay the wholesale price directly in each 30-minute block, plus the same network and environmental charges everyone pays, plus a flat $25 monthly subscription. Amber makes no margin on the energy. Their app shows the current price and a 24-hour forecast, and that forecast is the actual product.
| What you pay for energy | The 30-minute wholesale price, at cost |
|---|---|
| What you earn for exports | The same wholesale price, at cost |
| Amber's revenue | $25/month subscription only |
| Network & supply charges | Passed through, same as any retailer |
| Price protection | A wholesale cap on usage; it does not cover network charges |
| Contract term | None. No exit fee |
| Available in | NSW, VIC, SA, ACT, south-east QLD |
The one-sentence version: Amber is not a cheaper electricity plan, it is a different risk position. You take on the market's volatility and keep the savings your retailer used to keep for carrying it.
The technical detail, for people who want it
Thirty-minute settlement, five-minute dispatch
The market dispatches on five-minute intervals and prices settle across thirty. In practice a price event can be over before you have finished reading the notification. Automation matters far more than attention, which is the point of the next part.
SmartShift, and the evidence it works
SmartShift is Amber's battery automation. It ingests the price forecast and decides when the battery charges, holds, discharges or exports. Its objective is your bill, not Amber's revenue — Amber earns the same $25 whatever your battery does, which is the structural reason to trust it more than a VPP whose operator profits from your export decisions.
The clearest evidence I have that it works is a number most reviews would never think to look for. Across 87 days I exported only 40.6 kWh while the wholesale price was negative — that is, while exporting was costing me money — and the total cost of those moments was 47 cents. Solar curtailment is quietly doing its job thousands of times a month.
Negative feed-in tariffs, and curtailment
When rooftop solar across the state exceeds demand, the wholesale price goes negative and generators pay to keep exporting. On a fixed feed-in tariff you never see this; your retailer pays you 4c regardless and absorbs it. On Amber you see it directly, which means at certain moments exporting solar costs you money. The two answers are the battery — store the midday surplus and sell it later — and curtailment, which stops the inverter exporting while the price is below zero. Between them, negative pricing has gone from a problem to a rounding error for me. For a household with a large array and no storage it remains a real cost, and it is the main reason I would not recommend Amber to that household.
The price cap, and what it does not cover
Amber caps the wholesale component of your usage, which limits how bad a genuinely extreme event can get. Read the scope carefully: it caps wholesale energy. Network charges, metering, environmental scheme costs and the daily supply charge sit outside it. On my August bill the network daily supply charge was $35.33 and the Amber subscription $23.16 — neither of which any amount of clever battery dispatch will reduce.
The API, and why it is the best part
Amber publishes a REST API returning current, actual and forecast 30-minute prices for your site, and there is a mature Home Assistant integration built on it. Every figure in this review was pulled from that API directly. That is the strongest recommendation I can give it: the data is genuinely yours, at full resolution, and you can build against it.
With the forecast as a trigger rather than a timer, the automations write themselves: run hot water when the next hour is under 5c, hold the pool pump for the cheapest four-hour window, pause EV charging before a forecast spike. None of it requires a battery. If you already run Home Assistant, the API alone is a reason to look at Amber.
The actual numbers
These are measured, not modelled. They cover 87 days from 22 June 2026 to 16 September 2026, taken from Amber's API across all three channels — general household import, controlled-load import and feed-in.
| Grid imported | 867.9 kWh, costing $122.87 |
|---|---|
| Average import price | 14.16c/kWh |
| Exported | 930.0 kWh, earning $108.81 |
| Average export price | 11.70c/kWh |
| Net energy cost, 87 days | $14.06 |
| Share of import on controlled load | 81% |
| Share of import in the 4–8pm peak | 4.2% |
| Best export price captured | 46.5c/kWh, 9pm on 30 July |
| Highest import price seen | 67.4c/kWh |
| Previous plan, usage | 25.45c/kWh flat (14.42c controlled load) |
| Previous plan, feed-in | 4c/kWh (8.2c on a smart window) |
Three things in that table are worth pulling out, because they are the review.
My average import price is roughly half my old flat rate. 14.16c against 25.45c. That is not clever trading; it is what electricity costs when nobody is adding a retail margin and you are not forced to buy it at 6pm.
Only 4.2% of my grid import happens between 4pm and 8pm. This is the battery's real job, and it is almost invisible in a bill. The expensive part of the day simply stops being something I participate in.
Over 87 days my energy was near enough to break even — $14.06 net across import and export. I am not claiming free power; the supply charge and subscription still arrive every month, and this window is winter. But the energy itself, the part Amber actually changes, came out at roughly nothing.
The thing I did not expect: hot water is the whole story
81% of everything I pull from the grid is hot water on a controlled-load circuit. My general household import across 87 days was under two kilowatt hours a day — the solar and battery cover essentially all of it. Everything else is a tank heating overnight on a circuit the battery does not serve.
That single fact reframes the whole plan for me. The biggest remaining saving in my house is not a smarter battery strategy or a better price forecast; it is moving hot water off an overnight controlled-load circuit and onto the middle of the day, when my own solar is free. I would not have found that without half-hourly data, and no fixed-rate retailer would ever have shown it to me.
Why the money did not come from price spikes
Most enthusiastic Amber write-ups, including the one I intended to write, are built around price events: the evening the wholesale price hits several dollars a kilowatt hour, you export a full battery into it, and you earn a month's bill in thirty minutes.
Across 87 days, I caught exactly zero of those. Not one 30-minute interval paid me above $1/kWh. Not one paid above 50c. My best export price for the entire period was 46.5c/kWh at 9pm on 30 July, and my exports averaged 11.70c.
There are two honest reasons, and they point in opposite directions.
- Season. This window is winter and early spring. Serious price events in New South Wales cluster in summer heatwaves, and I have not been through one with the battery in place. My summer months on Amber were before the battery existed, which is exactly the wrong way round.
- Most of my export is just midday solar. The battery fills, the sun keeps shining, and the surplus goes out at whatever the middle of the day is paying — which is very little. That is not a failure of the plan, but it is a long way from the arbitrage story, and it is the honest description of where my 930.0 kWh went.
I would rather say this plainly than imply an income I have not earned. If you are considering Amber because you have read about somebody clearing hundreds of dollars in a single February evening, understand that it is real, it is rare, and it is not what a normal quarter looks like. The dependable gain is the boring one: a much lower average import price and a much better feed-in rate, every single day.
What went wrong, because something always does
My experience has been positive and I would sign up again. That is not the same as saying nothing went wrong.
The 25% state-of-charge floor
SmartShift will not discharge my battery below 25%, which on a 48kWh stack means roughly 12kWh is permanently unavailable to me. There are sound reasons for a reserve, but it is a quarter of an expensive asset that I cannot use, and it is not something I understood clearly when I bought the battery. If you are sizing a system around Amber, size it knowing that the top quarter of the number on the brochure is not yours.
The first three months were bad, and it was nobody's fault
I joined in January with no battery, in summer, with SmartShift still learning the house. That was my worst stretch by a distance. Amber is a plan that gets better as your hardware and its model of your house both improve, which is a genuinely awkward thing to sell — the experience you get in month one is not the one you are signing up for.
SmartShift occasionally makes a call I disagree with
Once or twice a month it does something I would not have: importing to charge when I would have waited, or discharging earlier in the evening than I would have chosen. Checking the forecast it was working from usually vindicates it. Sometimes I still think it was wrong. You can override it manually. This is the most common complaint in Australian Amber reviews and I do not think people are imagining it — they are correctly noticing that a system optimising against a forecast will sometimes lose, and that watching it lose on your own battery is more annoying than the dollar amount justifies.
Support is the weakest part of the product
I have not personally had a bad support experience, and I will not overclaim from a sample of one. But the volume of complaints about slow email responses in public reviews is large and consistent enough that leaving it out would mislead you. If you want a phone number answered quickly when a bill looks wrong, weigh that seriously.
The first month is genuinely unsettling
Watching a live price move while your house draws power is a different experience from getting a bill each quarter. For a few weeks I checked the app far too often. That passes, and the automation means it does not actually matter — but if you will find it stressful rather than interesting, that is a real cost that appears in nobody's savings calculation.
Who Amber Electric suits, and who it does not
Amber is likely a good fit if
- You have a home battery, of any size, and especially a large one.
- You have solar and can run big loads in the middle of the day.
- You charge an EV and can be flexible about when.
- You automate things, or want to — the API is excellent.
- You can absorb a bad month without it hurting.
- You will give it a year rather than a quarter.
Amber is likely a poor fit if
- You need the same number on every bill to budget.
- Your usage sits in the evening peak and cannot move.
- You have a large solar array and no storage.
- You are on a fixed or low income where one bad month is a crisis.
- You would rather never think about electricity pricing again.
- You want a retailer with fast, human phone support.
If you are in the right-hand column, no referral credit is a good enough reason to switch, including mine. The $120 is worth roughly two months of an average bill; the plan determines what you pay for years.
Amber versus a standard retailer, honestly compared
The comparison people usually make — Amber's rates against a competitor's rates — is the wrong one, because Amber does not have rates in that sense. Here is the comparison that actually decides it, with my own two plans as the example.
| Energy price | Amber: wholesale, at cost, changing every 30 minutes — mine averaged 14.16c. Previous plan: 25.45c flat, all day. |
|---|---|
| Feed-in tariff | Amber: wholesale — mine averaged 11.70c. Previous plan: 4c flat. |
| Fixed fee | Amber: $25/month subscription. Previous plan: none. |
| Who carries the risk | Amber: you. Previous plan: the retailer, and you pay them for it. |
| Bill predictability | Amber: low. Previous plan: high. |
| Upside from a battery | Amber: large. Previous plan: limited to self-consumption. |
| Visibility of your own data | Amber: full 30-minute history via API. Previous plan: a quarterly PDF. |
Read down the "who carries the risk" row and you have the whole decision. Everything else follows from it.
Before switching to anything, run your own numbers on Energy Made Easy, the free comparison service run by the Australian Energy Regulator. It lists every plan available at your address, including the ones that pay no commission and therefore never appear on commercial comparison sites.
The Amber Electric referral code, and what it is worth
If this review has been useful and you decide to join, you can use my referral code. I would rather say this in the middle of the page than hide it: I am paid $120 if you use it, on exactly the same terms you are. Amber pays both sides; your credit is not reduced by mine, and it is identical to the credit any other valid code would give.
$120 off
Applied as a $10 credit on each of your first 12 bills, not as a lump sum.
Sign up with this codeThe link fills the code in for you. Copying it and typing it into the referral field yourself gives an identical result.
Two honest notes. The credit is paid monthly across a year, so leaving after four months gets you $40, not $120. And $10 a month is most of Amber's $25 subscription — which is the accurate way to describe the offer: it is a discount on Amber's fee for your first year, not a discount on power. Full details are on the Amber Electric referral code page, which is kept current as Amber rotates codes.
What I would tell someone signing up tomorrow
- Find out what your controlled-load circuit is doing. Hot water was 81% of my grid import and I had no idea. If yours is similar, that is where your money is going, and it is fixable.
- Check your 4pm–8pm usage before anything else. That is the number Amber will charge you most for, and the number a battery is really buying down.
- Be realistic about storage. No battery and a large array? Model it carefully or wait. This is the single biggest predictor of whether you will be happy.
- Size a battery knowing you lose the bottom quarter to SmartShift's state-of-charge floor.
- Sign up with a referral code. They are free and the credit is real. Mine is FPKHAHAD.
- Expect a delay. Transfer takes time and a meter exchange takes more. Your credit starts at your first full bill.
- Turn the automation on and leave it alone for a month. You will make worse decisions than SmartShift does, and you will not enjoy yourself.
- Judge it over a year, not a quarter. My first three months were the worst and my most recent are the best. A season tells you something; a fortnight tells you nothing.
Common questions about Amber Electric
- Is Amber Electric worth it?
- For a household with solar and a battery that can shift load, yes — that has been my experience. Amber does not sell cheaper electricity; it sells wholesale exposure for a $25 monthly subscription. Over 87 measured days my grid import averaged 14.16c/kWh against the 25.45c flat rate I was paying before, and my exports averaged 11.70c/kWh against a 4c feed-in tariff. Without a battery, or without the ability to move usage, the same plan can easily cost more than a fixed rate.
- How much does Amber Electric cost per month?
- Amber charges a $25 monthly subscription on top of your usage, your daily supply charge and network costs. Amber adds no margin to the energy itself — you pay the 30-minute wholesale price, plus network, metering and environmental charges at cost. On my August bill the subscription was $23.16 before GST, and the network daily supply charge was actually the larger line at $35.33 for the month.
- Do you need a battery to use Amber Electric?
- No, but it changes the outcome more than anything else. My household import in the evening peak between 4pm and 8pm was 4.2% of my total grid import — the battery covers that window almost entirely. Without storage you are exposed to those prices with nothing to draw on, and your solar exports land in the middle of the day when wholesale prices are lowest and sometimes negative.
- Can your Amber bill spike?
- Yes, though it did not happen to me in the period I measured. The highest import price I saw across 87 days was 67.4c/kWh, and I imported almost nothing at it. That measured window is winter and early spring; summer is when the serious price events occur, and I have not yet been through one with the battery in place. Amber applies a wholesale price cap on usage, but it does not cover network, metering or environmental charges.
- What is SmartShift and does it work?
- SmartShift is Amber's battery automation. It forecasts wholesale prices and decides when the battery charges, holds, discharges or exports. It works well — the clearest evidence is that I exported only 40.6 kWh while the price was negative across 87 days, which cost me 47 cents in total. The curtailment is doing its job. My main complaint is the 25% state-of-charge floor it enforces, which keeps roughly 12kWh of my battery permanently out of play.
- Why is my Amber feed-in tariff negative?
- Because the wholesale price itself is negative. In the middle of a sunny day there is more rooftop solar on the grid than demand, and the market price falls below zero — generators pay to keep exporting. On Amber you see that directly, so exporting at that moment costs you money instead of earning it. A battery or solar curtailment fixes it: store the energy, or stop exporting until the price recovers.
- Which states is Amber Electric available in?
- New South Wales, Victoria, South Australia, the ACT and south-east Queensland. Amber does not operate in Western Australia, Tasmania or the Northern Territory, because those regions are not part of the National Electricity Market that Amber's pricing is built on.
- Do you need a smart meter for Amber?
- Yes. Amber bills against the 30-minute wholesale price, and an older accumulation meter cannot record consumption at that resolution. If your property does not have a smart meter, Amber arranges the upgrade as part of the switch, which adds a few weeks to the process.
- Is there an Amber Electric referral code?
- Yes. The current code is FPKHAHAD, verified on a live sign-up form on 16 September 2026. It gives you $120 off your bill, applied as a $10 credit on each of your first 12 bills rather than as a lump sum. The person whose code you use receives the same $120 on the same terms — including me, if you use mine, which is worth saying plainly rather than burying in a footer.
- Does Amber lock you into a contract?
- No. There is no fixed term and no exit fee, and you can leave at any time. That matters more than it sounds: trying Amber is a reversible decision. If three months of real bills show it does not suit your household, you switch back having lost the subscription fees and nothing else.
- Can you use Amber with Home Assistant?
- Yes. Amber publishes a REST API returning current, actual and forecast 30-minute prices for your site, and there is a well-maintained community integration for Home Assistant. Every figure in this review was pulled from that API rather than typed out of the app, which is a fair summary of how useful it is.
- Is Amber better than a standard retailer for solar?
- It depends entirely on whether you can move energy through time. A standard retailer pays a flat feed-in tariff regardless of what the energy is worth at that moment — mine was 4c/kWh. Amber pays what it is actually worth, which is close to nothing at midday and considerably more in the evening; mine averaged 11.70c/kWh. If you have storage, the second arrangement is better. If you export a lot at midday with nowhere to put it, the first one is.
The verdict
I am staying. My average import price is about half what a retailer charged me for the same electricity, my exports earn roughly three times what they used to, and the battery has taken my household almost entirely out of the evening peak. The API has made the house genuinely more interesting to run, and it found me a problem — an overnight hot water circuit — that no fixed-rate plan would ever have revealed.
But I would not have written that with a 10kWh battery, and I would not write it at all without one. I would also not have written it three months in. Amber rewards a specific kind of household with specific hardware, and it quietly penalises everyone else while charging them $25 a month. Work out which one you are before you switch, not after.
Figures in this review are my own, taken from Amber's API for the period 22 June 2026 to 16 September 2026, and from my own bills. Your results will differ. Nothing here is financial or energy advice, and Amber's pricing, offers and features change — confirm current terms with Amber before switching. Last updated 16 September 2026.