Tesla says it is almost sold out of Megapacks for 2027, and the Tesla Megapack 2027 order book is now one of the clearest hardware signals from the company’s Q3 2026 results. Speaking on the earnings call, CEO Elon Musk said, “Energy demand is very strong. We’re actually almost sold out of all Megapacks for 2027, and we’re ramping production as quickly as possible to meet demand.” Outlets including Yahoo Finance, Benzinga, CarSauce and Futu’s 富途牛牛 platform carried the same line.
We already covered the headline in our note on the Megapack near-sellout and its 110 GWh plan. This piece asks a different question. How does a 13.7 GWh quarter compare with roughly 110 GWh of planned factory capacity, and what can the word “almost” actually support?
Every number below comes from Tesla’s Q3 2026 materials and the reports on them. Where we do arithmetic, we label it as ours. Where the record is thin, we say so. Tesla has not published a percentage of 2027 capacity sold, so no one outside the company can verify one.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What Tesla Said About 2027 Megapack Orders
Three statements carry the story. The first is Musk’s own: Tesla is “almost sold out” of Megapacks for 2027. TeslaNorth reported the short form of that line as “We’re actually almost sold out of all Megapacks for 2027.” The second comes from Tesla’s own Q3 materials, which described the energy-storage business as having its “highest-ever order rate and backlog.” The third is CFO Vaibhav Taneja’s description of data-center demand for Megapacks as “extremely strong.”
Read together, those three phrases describe a business where orders are arriving faster than Tesla has ever booked them. They do not describe a business that has stopped taking orders. The gap between “almost” and “all” matters, and we return to it in a later section.
Musk also told investors that Tesla expects Megapack gross margins to “increase meaningfully.” He gave no number and no timeline in the reports we reviewed. That is typical for Tesla’s energy commentary, which tends to arrive as direction rather than data.
The Verified Numbers Behind the Tesla Megapack 2027 Story
Before testing any theory, it helps to list what is actually on the record. The table below holds only figures that Tesla’s Q3 2026 materials or the reports on them support. Nothing here is estimated.
| Metric | Reported figure | What it describes |
|---|---|---|
| Q3 2026 storage deployments | 13.7 GWh | Energy delivered to customers in one quarter |
| Megafactory Texas | 50 GWh per year (planned) | Production began in Q3 2026 |
| California capacity | 40 GWh per year (planned) | Existing Megapack manufacturing |
| Shanghai capacity | 20 GWh per year (planned) | Existing Megapack manufacturing |
| Combined planned capacity | About 110 GWh per year | Texas, California and Shanghai together |
| Apollo Power order | 260 Megapack 2XL units, about 1 GWh | Deliveries extend through the end of 2027 |
| Order rate and backlog | “Highest-ever” | Tesla’s own description of its Q3 position |
Two details deserve attention. First, the 110 GWh total is a sum of planned capacities, not a measure of what Tesla can build today. Second, the 13.7 GWh figure measures deployments, meaning energy that reached customers, not factory output. The two numbers describe different things, which is exactly why comparing them is useful and also why it needs care.
Megapack Sellout Math: 13.7 GWh vs 110 GWh
The sellout claim only matters if you know how much product exists to sell. That is where the gap between current deployments and planned capacity becomes the interesting part of the story.
The Annualized-Pace Check
Multiply the Q3 deployment figure by four and you get 54.8 GWh. That is our arithmetic, and it assumes every quarter matches Q3, which real quarters rarely do. Still, it gives a rough sense of the current run rate. Tesla’s planned capacity of about 110 GWh is roughly double that pace.
| Calculation (our arithmetic) | Inputs | Result |
|---|---|---|
| Annualized Q3 deployment pace | 13.7 GWh x 4 | 54.8 GWh |
| Planned capacity vs annualized pace | 110 GWh / 54.8 GWh | About 2.0x |
| Texas share of planned capacity | 50 GWh / 110 GWh | About 45% |
| California plus Shanghai planned | 40 GWh + 20 GWh | 60 GWh |
| Annualized pace vs pre-Texas capacity | 54.8 GWh / 60 GWh | About 91% |
| Apollo order vs one quarter | 1 GWh / 13.7 GWh | About 7% |
| Apollo order vs annual planned capacity | 1 GWh / 110 GWh | Under 1% |
| Average size per Apollo unit | About 1,000 MWh / 260 units | About 3.8 MWh |
The fifth row is the telling one. California and Shanghai together carry 60 GWh of planned capacity. Texas only began producing in Q3, so most of Q3’s output almost certainly came from those older sites. If so, they are already running close to their planned limits, and the sellout talk reflects a real squeeze on existing lines, not just marketing.
What the Ratio Cannot Tell You
The 2.0x ratio does not mean Tesla can double output next year. Planned capacity is a nameplate figure. Real production depends on cell supply, equipment yields, labor and how fast a new factory climbs its learning curve. Deployments also lag production because units ship, install and commission before they count. And Tesla has not said how much of the 13.7 GWh came from which site.
So treat the ratio as a ceiling on ambition, not a forecast. It shows how much room Tesla has built into its plan if everything works. It says nothing about whether everything will work by 2027.
Why “Almost Sold Out” Is Not the Same as Sold Out
Headlines have compressed Musk’s words into “sold out through 2027.” The on-record wording is narrower. Tesla said “almost” sold out, and the sources we reviewed never establish a percentage. We cannot tell you whether that means 90% of 2027 slots, 98%, or something else.
That precision matters for three reasons. A buyer planning a 2027 project needs to know whether any slots remain. An investor modeling energy revenue needs to know how much of next year is locked in. And a competitor deciding whether to chase Tesla’s customers needs to know whether those customers are already committed.
Tesla also has not said what “sold” means in this context. A signed purchase agreement, a deposit and a verbal commitment carry very different weight. The sources we reviewed do not define the term, and we will not guess. What we can say is that management chose a hedged word on a call where it had every incentive to sound as strong as possible. That restraint is itself a data point.
Megafactory Texas and the 50 GWh Ramp
Megafactory Texas began production in Q3 2026 and has a planned annual capacity of 50 GWh. On its own, that is nearly as large as the entire annualized Q3 deployment pace of 54.8 GWh. If Texas hits its plan, Tesla’s capacity base changes shape.
Musk’s phrasing on the call matters here. As The Tesla Space newsletter relayed it, Tesla is “ramping production as quickly as possible to meet demand.” That sentence concedes that production, not demand, is the limiting factor. Companies with spare capacity do not talk that way.
The timing risk is real. Tesla appears to have sold most of 2027 while the Texas line is still in its first quarter of production. That is a bet that the ramp goes to plan. If it slips by a few months, some 2027 deliveries slide into 2028 and customers feel it first. If it runs ahead, Tesla gains room to take more orders and the “almost” in “almost sold out” may quietly disappear.
Nevada LFP Megapacks and the Cell Supply Question
Tesla shipped its first Megapacks built exclusively with LFP (lithium iron phosphate) cells from its Nevada factory. For storage, that is a meaningful step. Cells make up a large share of a battery system’s cost and a larger share of its supply risk, so bringing them in-house shifts control from suppliers to Tesla.
The sources we reviewed do not quantify how many Megapacks will use Nevada cells or how much the change helps margins. We will not fill that gap with a guess. The same chemistry question showed up in our look at the Cybercab pack designed for an LFP swap, which suggests LFP is becoming a company-wide theme at Tesla rather than a one-product experiment.
For readers following grid hardware, the practical point is simple. A sellout that depends on third-party cells is exposed to someone else’s factory. A sellout that depends on in-house cells is exposed to Tesla’s own ramp. Neither is risk-free, but the second is easier for Tesla to manage and easier for outsiders to monitor.
Apollo Power’s 260-Unit Megapack 2XL Order in Context
Apollo Power agreed to buy 260 Megapack 2XL units, reported at approximately 1 GWh, with deliveries extending through the end of 2027. It is the one named customer order in the reports we reviewed, which makes it a useful test case.
Our arithmetic puts the order at about 7% of a single Q3 and under 1% of the 110 GWh planned annual capacity. Spread across 260 units, it averages roughly 3.8 MWh each. Those are modest slices. The takeaway is not that Apollo alone drains Tesla’s supply. It is that a sellout on this scale is built from many orders of this size, and Tesla has disclosed only one.
Delivery timing is the other lesson. An order whose deliveries run through the end of 2027 occupies Tesla’s calendar for the whole year, not a single quarter. That is how a company can be nearly sold out for 2027 while Q4 2026 is still in progress. Long, staggered delivery schedules turn a handful of orders into a full-year commitment.
Data Center Demand and the Storage Queue
Taneja called demand for Megapacks from data centers “extremely strong.” That fits a wider pattern in hardware, where AI buildouts are stretching every part of the supply chain, from accelerators to power equipment.
Large data-center campuses need dependable power and have limited patience for grid delays. On-site storage can smooth peaks and add flexibility while a campus waits for utility interconnection. That is a general industry logic, not a Tesla disclosure. The reports we reviewed do not break out how much of the backlog comes from data centers versus utilities or developers, so we cannot say what share this one segment holds.
What the record does show is direction. A CFO singling out one customer segment as extremely strong, in the same quarter as a record backlog, tells you where incremental orders are coming from. If that segment keeps growing, it can push 2028 slots into the same queue before 2027 even begins.
Margin Outlook: What “Increase Meaningfully” Leaves Out
Musk said Megapack gross margins should “increase meaningfully.” Investors will want to know from what base, by how much and by when. None of that is in the sources we reviewed, and Tesla’s materials as described in reports do not supply a number.
There are plausible drivers. Scale from a new factory spreads fixed costs. In-house LFP cells reduce dependence on outside suppliers. A tight order book gives Tesla pricing leverage. But these are reasons margins could rise, not proof that they will, and the sources do not tell us which carries the most weight.
Readers should also keep company-wide profit in view. Tesla’s Q3 gross profit already carried a one-off lift, as we explained in our breakdown of tariff refunds and gross profit. Separating that effect from underlying storage economics is the first job for anyone modeling the segment.
Market Impact for Investors, Rivals and Grid Buyers
For investors, the near-sellout turns energy storage from a side story into a driver of next year’s numbers. Analysts were already leaning on the segment. Our report on how Barclays tied its Tesla upgrade to energy shows how much of the bull case rests on it. A sellout claim backs that thesis with an order-book signal, though one that is not yet quantified.
For rivals, the message cuts both ways. A Tesla that cannot supply every buyer leaves demand for other integrators and cell makers. A Tesla that is adding 50 GWh in Texas signals it intends to take that demand back. We did not verify rival backlog or capacity figures in this run, so we describe the dynamic and leave the competitor numbers to their own disclosures.
For grid buyers and developers, the impact is practical. Procurement windows are narrowing. Anyone who has not locked a 2027 slot should assume Tesla’s calendar is close to full and plan for either a later delivery date or a different vendor.
Historical Context: From Side Business to Backlog Story
For years, Tesla’s energy segment trailed far behind vehicles in investor attention. Cars set the headlines and storage was the footnote. The Q3 2026 language reverses that order. Tesla’s own materials now describe a record order rate and backlog, and management is building capacity measured in tens of gigawatt-hours across three sites.
The pattern is familiar from other AI-era hardware shortages. Demand shows up first as a vague signal from management. Capacity announcements follow. Pricing pressure comes last. We saw a similar sequence in storage chips, where analysts now expect enterprise SSD prices to rise in Q4 as suppliers race to meet AI demand. Megapacks sit earlier in that sequence, with capacity being built and pricing still undisclosed.
That earlier position is the reason to watch the next two or three quarters closely. If deployments rise as Texas ramps, the sellout reads as demand meeting new supply. If deployments stall, it reads as a bottleneck. Our hardware coverage tracks the same supply-and-demand tension across chips, storage and power.
Competitive Comparison: How to Read a Sellout Claim
Sellout statements are easy to make and hard to compare. The table below sets out what Tesla’s disclosures confirm against what stays unconfirmed. Use the same checklist on any storage vendor that says its order book is full.
| Question a buyer should ask | What is confirmed | What is unconfirmed |
|---|---|---|
| How much of 2027 is sold? | “Almost all,” per Tesla management | Any exact percentage |
| Is 2027 completely sold out? | Not claimed on the record | A full sellout through all of 2027 |
| Is the backlog growing? | Highest-ever order rate and backlog | Backlog value in dollars |
| Is new capacity coming? | Texas started in Q3, 50 GWh planned | Actual output by quarter |
| What do systems cost? | Nothing on the record | Megapack prices and contract values |
| Who is buying? | Apollo Power, 260 Megapack 2XL units | Customer-by-customer backlog |
| Were earlier years sold out? | Nothing on the record | Any sellout before 2027 |
The unconfirmed column is longer than the confirmed one, and that is the honest state of the public record. Tesla gave direction, a few capacity numbers and one named order. It did not give prices, dollar backlog or a sold percentage. Any rival that publishes all three would be giving buyers more than Tesla has.
What Buyers and Developers Should Do Now
If you plan to buy utility-scale storage for 2027, start by asking Tesla’s sales team for a written delivery window, not a general assurance. “Almost sold out” leaves open the possibility of remaining slots, but those slots may fall late in the year.
Next, build flexibility into the schedule. Staggered deliveries like Apollo’s suggest Tesla can fit large orders into a long calendar. Splitting your own order across quarters may improve your odds of securing capacity. Also ask which cell chemistry your units will use, since Tesla has started shipping Megapacks with Nevada-made LFP cells and your contract should state what you receive.
Finally, keep a second vendor in the conversation. A near-sold-out supplier has little reason to discount, and a delay at a single new factory could ripple through every customer in the queue. A backup quote costs little and protects the project timeline.
Five Predictions for Megapack Through 2027
These are our expectations, built from the record above. They are analysis, not facts, and any of them could miss.
- Tesla keeps hedged language. Management will keep saying “almost” or “nearly” and will not publish a sold percentage in the next two quarters.
- Quarterly deployments climb past 13.7 GWh. Texas output will show up in deployment figures gradually, with the biggest effect landing in 2027 rather than late 2026.
- Margin talk stays qualitative. Tesla will point to Texas scale and Nevada LFP cells without giving a segment margin target in numbers.
- 2028 slots enter the conversation. As 2027 fills, buyers will push for 2028 commitments, and management will start describing that year’s order book.
- More named customers surface. Apollo Power will not be the only buyer Tesla or its customers disclose, especially if data-center demand stays as strong as the CFO described.
Frequently Asked Questions
Is Tesla completely sold out of Megapacks for 2027? No. Elon Musk said Tesla is “almost sold out,” and the sources we reviewed do not confirm a full sellout. The exact share of 2027 capacity that remains is unconfirmed.
How much energy storage did Tesla deploy in Q3 2026? Tesla reported 13.7 GWh of energy-storage deployments in Q3 2026. Its materials also described the business as having its highest-ever order rate and backlog.
How big is Megafactory Texas? Megafactory Texas began production in Q3 2026 and has a planned annual capacity of 50 GWh. Reported planned capacities of 40 GWh in California and 20 GWh in Shanghai bring the total to approximately 110 GWh.
What is the Apollo Power order? Apollo Power agreed to buy 260 Megapack 2XL units, reported at approximately 1 GWh. Deliveries extend through the end of 2027.
What changed with Megapack batteries? Tesla shipped its first Megapacks built exclusively with LFP cells from its Nevada factory. The sources we reviewed do not say how much the change adds to margins.
Did Tesla give Megapack prices or contract values? No. Megapack prices, total contract value and customer-by-customer backlog are unconfirmed in the reports we reviewed.
Will Megapack margins rise? Musk said Tesla expects Megapack gross margins to “increase meaningfully.” He gave no figure or date in the reports we reviewed, so the size and timing of any increase remain open.
