# META Full forecast · Q4 2026

> The forecast for Meta: the next two quarters line by line, three years of sales and EPS, a low and a high case. Next report: Q4 2026 in January 2027. Date not yet set.

Canonical: https://pricedin.markets/meta/fq4-2026/forecast/

cutoff 21 October 2026, 17:31 ET

The forecast for Meta: the next two quarters line by line, three years of sales and EPS, a low and a high case. Next report: Q4 2026 in January 2027. Date not yet set.

[Quarter forecast →](https://pricedin.markets/meta/fq4-2026/preview/)

META · as of 2026-10-21 17:31:12 ET · 6 AI forecasters · 80% ranges, 1 in 10 to 9 in 10

## Sales and earnings outlook

Next quarter · Q4 2026

Sales $75.0B inside the $73.0B–$76.0B outlook

Adjusted EPS $8.23 −7% on Q4 2025

Quarter after · Q1 2027

Sales $68.6B +22% on Q1 2026

Adjusted EPS $6.13 −41% on Q1 2026

This year · FY2026

Sales $257B +28% on FY2025

Adjusted EPS $32.24

Next year · FY2027

Sales $309B +20% on FY2026

Adjusted EPS $31.42 −3% on FY2026

Year +2 · FY2028

Sales $361B +17% on FY2027

Adjusted EPS $37.05 +18% on FY2027

##### Sales

PricedIn **$75.0B** $72.2B – $77.4B

$72.2B $75.0B $77.4B

##### adjusted EPS

PricedIn **$8.23** $6.87 – $9.57

$6.87 $8.23 $9.57

**Middle** the forecast **80% range** 1 in 10 chance below this bar, 1 in 10 above it. These are the low and high printed by each number.

The models agree closely: sales middles run from $74.5B to $75.6B. Muse, lowest, placed its middle 'at 74.5 billion, inside that range' without leaning high. Opus, highest, notes the first-quarter result 'landed near the top of this range' and puts the quarter near the top of the guide with costs 'a little below the midpoint'. Sol's earnings range, $6.10 to $10.50, is widest: its 'approximate 10th-to-90th-percentile sales range' is 'wider than management's planning range'.

Fourth-quarter sales of about $75.0B, in the upper half of the company's own range, as AI-driven ad gains carry into the holidays.

Sales **$75.0B** $72.2B – $77.4B · +25% on Q4 2025 ($59.9B)

Adjusted EPS **$8.23** $6.87 – $9.57 · −7% on Q4 2025 ($8.88)

#### Investment and cash generation

Base case

Capital spending **$52.9B**

Free cash flow **−$14.0B**

Free cash flow: operating cash flow less capital spending.

The company reported third-quarter sales of $64.6B, above the top of its range and 2.3% above our forecast. Ad prices rose 14% and ad impressions 10% from a year earlier. A $9.9B charge for the state attorneys general settlement cut reported operating margin to 20%; without it, earnings per share would have been $7.40.

For the fourth quarter the company guides sales of $73B to $76B, with currency about a 1% drag, and full-year expenses of $175B to $179B. It expects a tax rate of 15% to 17% for the quarter.

All six models put the quarter's sales between $74.5B and $75.6B, an average of $75.0B, about 16% above the third quarter. Most lean to the upper half of the range because 2026 quarters have landed near or above the top of the company's ranges.

Costs implied by the full-year guide give an operating margin near 34% and earnings per share of about $8.23. Capital spending near $52.9B leaves free cash flow around −$14.0B for the quarter.

The models differ more on earnings than on sales. Sol and Grok let the quarter miss or beat the company's range; GLM and Opus keep their ranges close to it.

About $8.23 per share: sales near $75.0B, costs implied by the full-year expense guide, a tax rate near 16%, and today's share count before the planned share dividend.

#### Outlook against result

Q2 2025 +4.4% above $42.5B–$45.5B → $47.5B guided “$42.5-45.5 billion”

Q3 2025 +1.5% above $47.5B–$50.5B → $51.2B guided “$47.5-50.5 billion”

Q4 2025 +1.5% above $56.0B–$59.0B → $59.9B guided “$56-59 billion”

Q1 2026 inside $53.5B–$56.5B → $56.3B guided “$53.5-56.5 billion”

Q2 2026 inside $58.0B–$61.0B → $60.8B guided “$58-61 billion”

Q3 2026 +0.9% above $61.0B–$64.0B → $64.6B guided “$61-64 billion”

Q4 2026 inside $73.0B–$76.0B → $75.0B guided “$73-76 billion”

Each outlook is the one the company gave before the quarter. The last row is the models' forecast.

#### The quarter forecast, line by line

| Line | Q3 2026 reported | Q4 2025 reported | Q4 2026 forecast | Q1 2026 reported | Q1 2027 forecast |
| --- | --- | --- | --- | --- | --- |
| Revenue | $64.6B | $59.9B | $75.0B | $56.3B | $68.6B |
| Sales growth, year on year | +26% | +24% | +25% | +33% | +22% |
| Family of Apps | $64.2B | $58.9B | $74.0B | $55.9B | $68.2B |
| Reality Labs | $443M | $955M | $983M | $402M | $468M |
| Total costs and expenses | $51.6B | $35.1B | $49.7B | $33.4B | $49.6B |
| Income (loss) from operations | $13.0B | $24.7B | $25.3B | $22.9B | $19.0B |
| Operating margin | 20.1% | 41.3% | 33.7% | 40.6% | 27.7% |
| Interest and other income (expense), net | $0.01B | $0.61B | −$0.01B | −$1.12B | −$0.15B |
| Income before income taxes | $13.0B | $25.4B | $25.3B | $21.8B | $18.9B |
| Provision (benefit) for income taxes | $3.14B | $2.59B | $4.04B | −$5.02B | $3.00B |
| Tax rate | 24.2% | 10.2% | 16.0% | −23.1% | 15.9% |
| Net income | $9.9B | $22.8B | $21.2B | $26.8B | $15.9B |
| Diluted (in shares) | 2.57B | 2.56B | 2.58B | 2.56B | 2.59B |
| Adjusted EPS | — | $8.88 | $8.23 | $10.44 | $6.13 |

The lines as the company's statement shows them. The forecast columns are the average of the models, line by line. Sales growth is the change on the same quarter a year earlier. Earnings per share is the headline's: the company's own for the reported quarter, and for a forecast quarter worked out from the averaged lines with the company's adjustments.

#### What it rests on

**Fourth-quarter sales outlook**

$73B to $76B, with currency an approximately 1% headwind “We expect fourth quarter 2026 total revenue to be in the range of $73-76 billion. Our guidance assumes foreign currency is an approximately 1% headwind to year-over-year total revenue growth, based on current exchange rates.” Third-quarter 2026 earnings release

**Full-year expense outlook**

$175B to $179B, raised to include the $9.9B legal charge “We are raising our expense outlook to incorporate the $9.9 billion charge related to legal proceedings recognized in the third quarter. We now expect full year 2026 total expenses to be in the range of $175-179 billion.” Third-quarter 2026 earnings release

**Fourth-quarter tax rate**

Between 15% and 17% “Absent any changes to our tax landscape, we expect our fourth quarter 2026 tax rate to be between 15-17%.” Third-quarter 2026 earnings release

**Third-quarter ad prices**

Average price per ad up 14% from a year earlier “•Average price per ad – Average price per ad increased by 14% year-over-year.” Third-quarter 2026 earnings release

**Third-quarter adjusted earnings**

$7.40 per share without the legal charge, against $3.83 reported “•Diluted EPS of $3.83 would have increased by $3.57 to $7.40.” Third-quarter 2026 earnings release

#### Low and high

Base the middle forecast

Key assumptions

- ****Sales land in the upper half of the $73B to $76B guide, about $75.0B.** 2026 quarters landed near or above the top of the company's ranges; third-quarter sales reached $64.6B. *4 of 6***
  - **Opus 5.5** I put fourth quarter revenue near the top, about 75.6 billion
  - **GPT-6.1 Sol** My middle is $75.1 billion, modestly above the midpoint
  - **Grok 4.7** the middle is set at 75 billion dollars, in the upper half
  - **Gemini 4 Argon** fourth quarter 2026 revenue reaches 75 billion dollars (near the upper half of management's 73 billion to 76 billion dollar guidance range)
- ****Costs near $49.7B, within what the full-year expense guide leaves for the quarter.** Nine-month costs were about $127B, so the guide leaves roughly $48B to $52B. *4 of 6***
  - **Opus 5.5** I used 49.3 billion, a little below the midpoint
  - **Grok 4.7** fourth-quarter costs land near 50 billion dollars at the midpoint
  - **GLM 5.3** the expense outlook implies Q4 costs of roughly $48–52B
  - **Muse Wonder** imply fourth quarter costs near 50 billion
- ****No new legal charge; the $9.9B settlement accrual stays in the third quarter.** Extra payments depend on YouTube and TikTok joining, which the company does not treat as probable. *4 of 6***
  - **Opus 5.5** I left that out of the middle
  - **Muse Wonder** I assumed no extra accrual in the base case
  - **Grok 4.7** so they are not in the sales or earnings middle
  - **GLM 5.3** neither Q4-2026 nor 2027 carries a repeat
- ****Tax rate of about 16%, the middle of the company's 15% to 17% guide.** The third quarter's 24% rate came from the non-deductible legal charge, which does not repeat. *3 of 6***
  - **Grok 4.7** The middle uses 16 percent.
  - **Opus 5.5** Tax 16%.
  - **Muse Wonder** The guided fourth quarter tax rate of fifteen to seventeen percent sets the quarter plan tax line.
- ****Capital spending near $52.9B, leaving free cash flow around −$14.0B.** Full-year spending of $135B to $145B leaves a very heavy final quarter. *3 of 6***
  - **GPT-6.1 Sol** Fourth-quarter purchases are about $52.9 billion.
  - **Muse Wonder** fourth quarter capital spending near 52 to 54 billion including finance leases
  - **GLM 5.3** implying a very heavy Q4

If all of this happens

**$75.0B** sales

**$8.23** adjusted EPS

**$52.9B** Capital spending

**−$14.0B** Free cash flow

Cash walk not supplied by every model.

Low about 1 chance in 10 of worse

What the low case assumes

- ****Holiday ad demand weakens under high energy prices and weak consumer confidence.** Advertiser spending dipped when the Iran conflict began, and oil prices remain very high. *3 of 6***
  - **Grok 4.7** Marketer budgets shrink as high energy prices and soft consumer confidence persist or worsen
  - **Opus 5.5** advertiser demand under high oil prices and weak consumer sentiment
  - **Gemini 4 Argon** Macroeconomic deceleration from high oil prices and elevated interest rates reducing digital advertising demand
- ****Sales miss the bottom of the $73B to $76B range.** Grok and Sol set their lows outside the guide as a one-in-ten miss. *3 of 6***
  - **Grok 4.7** The low and high sit outside the guide, as a 1-in-10 miss or beat.
  - **GPT-6.1 Sol** The approximate 10th-to-90th-percentile sales range is $71.5 billion to $78.5 billion, wider than management's planning range.
  - **Muse Wonder** low and high allowing a small miss or beat
- ****A new legal charge lands, which adjusted earnings would not exclude.** Only the named settlement accrual is excluded; youth-related trials are scheduled in coming months. *3 of 6***
  - **Opus 5.5** any new charge would fall in my low case
  - **Grok 4.7** a new legal accrual of several billion dollars is booked
  - **Muse Wonder** allowed for new legal charges in the low case

If all of this happens

**$72.2B** sales · −4% vs base

**$6.87** EPS · −17% vs base

Sales would have to fall to about $72.2B, 3.7% below the base, with earnings near $6.87. That needs holiday ad budgets to give way under high oil prices and weak consumer mood, or a new legal charge that adjusted earnings would not exclude. Grok goes furthest, to $70.8B and $6.00, because it lets sales miss the company's range outright while capital spending stays at the top of its guide.

Advertiser spending already dipped when the Iran conflict began in March 2026, yet third-quarter sales still grew 26%.

The six lows run from $70.8B (Grok 4.7) to $73.6B (Opus 5.5).

First sign A new legal accrual or softer holiday ad pricing disclosed before the quarter closes on December 31, 2026.

High about 1 chance in 10 of better

What the high case assumes

- ****Holiday ad demand stays as strong as in the third quarter.** Ad prices rose 14% in the third quarter on ad performance gains. *3 of 6***
  - **Grok 4.7** Holiday advertising would have to stay as strong as the third quarter
  - **Opus 5.5** sustained high-20s ad growth from better ranking and new surfaces
  - **Gemini 4 Argon** Accelerated advertising conversion lifts from generative retrieval models
- ****Costs land at the low end of the expense guide.** The May headcount cut left employees down 12% from the second quarter. *2 of 6***
  - **Grok 4.7** spending stays at the low end of the guide
  - **Opus 5.5** cost discipline after the headcount reduction
- ****Tax rate at the low end of the 15% to 17% guide.** The company guides that range absent changes to its tax landscape. *1 of 6***
  - **Grok 4.7** The tax rate stays at the low end of the guided range

If all of this happens

**$77.4B** sales · +3% vs base

**$9.57** EPS · +16% vs base

Sales would have to reach about $77.4B, 3.2% above the base, with earnings near $9.57. Holiday advertising would need to stay as strong as the third quarter, with costs and tax at the low end of the guides. Sol goes furthest, to $78.5B and $10.50, because it sets its high as a beat beyond the company's range.

Third-quarter sales of $64.6B beat the top of the company's $61B to $64B range.

The six highs run from $76.4B (GLM 5.3) to $78.5B (GPT-6.1 Sol).

First sign Holiday ad pricing through December 31, 2026, then sales above $76B at the next report.

#### Watch

Sales above $76B would point higher; sales at or below $73B, or a new legal charge, would point lower. The next report also brings the first-quarter outlook and the 2027 expense and capital spending plans.

##### Sales

PricedIn **$68.6B** $64.2B – $72.9B

$64.2B $68.6B $72.9B

##### adjusted EPS

PricedIn **$6.13** $4.70 – $7.65

$4.70 $6.13 $7.65

**Middle** the forecast **80% range** 1 in 10 chance below this bar, 1 in 10 above it. These are the low and high printed by each number.

The models agree on a seasonal dip, from 6.3% to 11% below the fourth quarter. Opus, lowest on sales at $67.6B, holds that 'the first quarter is seasonally the weakest for revenue while costs stay near fourth quarter levels.' Grok, highest at $70.2B, sees 'growth slows but stays near 20 percent in 2027'. On earnings Sol is lowest, $5.11, because 'faster model efficiency is not assumed to reduce total serving expense when usage grows.'

First-quarter 2027 sales of about $68.6B, the usual seasonal dip after the holiday quarter but well above a year earlier.

Sales **$68.6B** $64.2B – $72.9B · +22% on Q1 2026 ($56.3B)

Adjusted EPS **$6.13** $4.70 – $7.65 · −41% on Q1 2026 ($10.44)

#### Investment and cash generation

Capital spending: not supplied by every model

Free cash flow: not supplied by every model

Free cash flow: operating cash flow less capital spending.

The company has not yet given a first-quarter 2027 outlook; it will at its next report. A year earlier it guided $53.5B to $56.5B and reported $56.3B, near the top.

The models expect sales of about $68.6B, 8.5% below the fourth quarter, the normal pattern after holiday advertising. Their middles run from $67.6B for Opus to $70.2B for Grok.

Costs stay near fourth-quarter levels while sales dip, so operating margin falls to about 28% and earnings per share to about $6.13. A year earlier the company earned $10.44, lifted by a one-time tax benefit; without it, $7.31.

Earnings are where the models part. Sol is lowest at $5.11, seeing serving costs and depreciation rise fast; GLM is highest at $7.25, with a margin held up by narrowing Reality Labs losses and a smaller workforce.

The planned share dividend, which works like a two-for-one split, is not applied: all per-share figures stay on today's share count.

About $6.13 per share: sales near $68.6B, an operating margin near 28%, a tax rate near 16%, and no repeat of last year's first-quarter tax benefit.

#### The quarter forecast, line by line

| Line | Q3 2026 reported | Q4 2025 reported | Q4 2026 forecast | Q1 2026 reported | Q1 2027 forecast |
| --- | --- | --- | --- | --- | --- |
| Revenue | $64.6B | $59.9B | $75.0B | $56.3B | $68.6B |
| Sales growth, year on year | +26% | +24% | +25% | +33% | +22% |
| Family of Apps | $64.2B | $58.9B | $74.0B | $55.9B | $68.2B |
| Reality Labs | $443M | $955M | $983M | $402M | $468M |
| Total costs and expenses | $51.6B | $35.1B | $49.7B | $33.4B | $49.6B |
| Income (loss) from operations | $13.0B | $24.7B | $25.3B | $22.9B | $19.0B |
| Operating margin | 20.1% | 41.3% | 33.7% | 40.6% | 27.7% |
| Interest and other income (expense), net | $0.01B | $0.61B | −$0.01B | −$1.12B | −$0.15B |
| Income before income taxes | $13.0B | $25.4B | $25.3B | $21.8B | $18.9B |
| Provision (benefit) for income taxes | $3.14B | $2.59B | $4.04B | −$5.02B | $3.00B |
| Tax rate | 24.2% | 10.2% | 16.0% | −23.1% | 15.9% |
| Net income | $9.9B | $22.8B | $21.2B | $26.8B | $15.9B |
| Diluted (in shares) | 2.57B | 2.56B | 2.58B | 2.56B | 2.59B |
| Adjusted EPS | — | $8.88 | $8.23 | $10.44 | $6.13 |

The lines as the company's statement shows them. The forecast columns are the average of the models, line by line. Sales growth is the change on the same quarter a year earlier. Earnings per share is the headline's: the company's own for the reported quarter, and for a forecast quarter worked out from the averaged lines with the company's adjustments.

#### What it rests on

**Last year's first-quarter outlook**

$53.5B to $56.5B; the quarter came in at $56.3B “We expect our first quarter 2026 total revenue to be in the range of $53.5 billion to $56.5 billion.” Fourth-quarter 2025 earnings call

**Last year's tax benefit**

Without it, first-quarter 2026 earnings would have been $7.31 per share “Absent the tax benefit, our net income and EPS would have been $18.7 billion and $7.31, respectively.” First-quarter 2026 earnings call

**Muse serving costs**

Costly to run now; company expects over tenfold efficiency gain within a year “In the near term, Muse is very compute intensive to run, but we expect to improve its cost efficiency by more than 10x in the next year, which will enable it to be profitable once the revenue lines are at scale.” Third-quarter 2026 earnings call

**Capacity plans**

Plans geared to maximizing 2026 and 2027 computing capacity “Consequently, our current plans are geared towards maximizing 2026 and 2027 capacity.” Second-quarter 2026 earnings call

#### Low and high

Base the middle forecast

Key assumptions

- ****Sales dip about 8.5% from the fourth quarter, the normal seasonal pattern.** Holiday advertising peaks in the fourth quarter; the first quarter is usually the softest. *3 of 6***
  - **Opus 5.5** the first quarter is seasonally the weakest for revenue
  - **GLM 5.3** why Q1 typically dips 5–7% from Q4
  - **Gemini 4 Argon** an eight to twelve percent seasonal pullback in the first quarter
- ****Costs stay near fourth-quarter levels, so operating margin falls to about 28%.** Depreciation and cloud costs from the capacity build keep rising into 2027. *3 of 6***
  - **Opus 5.5** put the first quarter 2027 margin near 28%
  - **GPT-6.1 Sol** Larger model-serving costs and depreciation offset much of the advertising growth
  - **Muse Wonder** Costs stay very high as capacity built in 2026 and 2027 comes into use
- ****No repeat of the first-quarter 2026 tax benefit; tax rate near 16%.** The $8.03B benefit was one-time; without it first-quarter 2026 earnings were $7.31. *3 of 6***
  - **Grok 4.7** That benefit does not repeat
  - **GPT-6.1 Sol** The benefit does not recur in my 2027 or 2028 middle.
  - **GLM 5.3** about 16% for 2027–28
- ****Per-share figures stay on today's share count; the planned share dividend is not applied.** The stock dividend was not paid at the cutoff. *3 of 6***
  - **Opus 5.5** every per-share figure stays on today's roughly 2.57-2.62 billion diluted shares
  - **Grok 4.7** no share count and no earnings per share figure is split
  - **GLM 5.3** every share count and per-share figure in my forecast stays on the current (pre-split) share basis

If all of this happens

**$68.6B** sales

**$6.13** adjusted EPS

Low about 1 chance in 10 of worse

What the low case assumes

- ****Ad impressions and prices grow well below the recent trend.** Third-quarter impressions rose only 10%, and part of the 14% price gain came from easier comparisons. *3 of 6***
  - **Muse Wonder** Advertising impressions and price growth fall well below the recent trend
  - **GPT-6.1 Sol** Weaker ad pricing and engagement growth
  - **Opus 5.5** price per ad growth falls to low single digits
- ****Serving and depreciation costs rise faster than sales.** Data center leases not yet begun totaled about $278.99B in June 2026. *3 of 6***
  - **GPT-6.1 Sol** Computing costs, legal exposure and financing pressure
  - **Muse Wonder** Infrastructure and legal costs rise faster than revenue
  - **Gemini 4 Argon** High fixed depreciation and multi-year third-party cloud commitments weighing on operating margins
- ****High energy prices and weak consumer confidence cut advertiser budgets.** Advertiser spending dipped when the Iran conflict began in March 2026. *3 of 6***
  - **Opus 5.5** advertiser demand under high oil prices and weak consumer sentiment
  - **Grok 4.7** Marketer budgets shrink as high energy prices and soft consumer confidence persist or worsen
  - **Gemini 4 Argon** Macroeconomic deceleration from high oil prices and elevated interest rates reducing digital advertising demand

If all of this happens

**$64.2B** sales · −6% vs base

**$4.70** EPS · −23% vs base

Sales would have to fall to about $64.2B, 6.4% below the base, with earnings near $4.70. Ad impressions and prices would need to grow well below trend while serving and depreciation costs keep climbing. Muse goes furthest on sales, to $62.0B, with ad growth falling well below the recent trend; Sol goes furthest on earnings, to $3.40, because it requires cost and tax pressure on top of weak sales.

Ad impression growth already slowed from 19% to 14% to 10% over the first three quarters of 2026.

The six lows run from $62.0B (Muse Wonder) to $66.5B (GLM 5.3).

First sign A first-quarter 2027 outlook below about $68B at the company's next report, whose date is not yet given.

High about 1 chance in 10 of better

What the high case assumes

- ****Ad impressions and prices both keep growing at double-digit rates.** Third-quarter ad prices rose 14% and impressions 10%. *3 of 6***
  - **Grok 4.7** Ad impressions and prices both keep growing at double-digit rates through 2027
  - **Opus 5.5** impressions low-teens, price per ad low-to-mid teens
  - **GPT-6.1 Sol** Sustained strong advertising conversions and pricing
- ****Muse, business agents and subscriptions bring paid revenue sooner.** Over 10 million people set up Muse in six weeks; subscriptions and trials passed 15 million. *3 of 6***
  - **GLM 5.3** Muse scaling from 10 million setups toward a paid, commerce-monetized network
  - **Muse Wonder** Muse, business agents and model services become material paid revenue in 2027 and 2028
  - **Gemini 4 Argon** Fast global scaling and high-margin commerce and subscription monetization of the Muse personal agent
- ****Compute efficiency improves sharply, holding costs down.** The company expects more than tenfold cost-efficiency gains over the next year. *3 of 6***
  - **GLM 5.3** Compute cost efficiency improving more than tenfold as promised
  - **Muse Wonder** Compute efficiency improves more than tenfold
  - **Gemini 4 Argon** Tenfold improvements in artificial intelligence inference efficiency

If all of this happens

**$72.9B** sales · +6% vs base

**$7.65** EPS · +25% vs base

Sales would have to reach about $72.9B, 6.2% above the base, with earnings near $7.65. Ad growth would need to hold near holiday strength into the new year, paid products would add revenue, and compute efficiency would hold costs down. Grok goes furthest on sales, to $74.2B, because ad impressions and prices both keep growing at double-digit rates; Grok and GLM reach $8.30 on earnings.

The first quarter of 2026 fell 6.0% from the fourth quarter of 2025; a shallower dip would point to the high case.

The six highs run from $71.0B (Opus 5.5) to $74.2B (Grok 4.7).

First sign A first-quarter 2027 outlook implying growth of 26% or more at the company's next report.

#### Watch

The first-quarter outlook at the next report: a range centered above $68.6B points higher, below about $68B points lower. Also watch the 2027 expense plan and any paid revenue from Muse.

##### Sales

PricedIn **$257B** $254B – $259B

$254B $257B $259B

##### adjusted EPS

PricedIn **$32.24** $30.75 – $33.73

$30.75 $32.24 $33.73

**Middle** the forecast **80% range** 1 in 10 chance below this bar, 1 in 10 above it. These are the low and high printed by each number.

All six land between $256B and $257B on sales and $32.10 to $32.50 on earnings. GLM and Muse, lowest, put fourth-quarter sales at or under the guide's midpoint; GLM uses a 'revenue midpoint $74.5B'. Opus, highest at $32.50, puts sales near the top of the guide and costs 'a little below the midpoint'. Grok's earnings range, $29.50 to $34.80, is the widest.

2026 sales of about $257B, with the holiday quarter capping a year of AI-driven ad gains.

Sales **$257B** $254B – $259B · +28% on FY2025

Adjusted EPS **$32.24** $30.75 – $33.73

Base case

Operating margin **31.1%** 41.4% in FY2025

#### Investment and cash generation

Base case

Capital spending **$137B**

Free cash flow **−$1.04B**

Free cash flow: operating cash flow less capital spending.

Nine months have brought in $182B of sales. With the fourth quarter near $75.0B, the year reaches about $257B, 28% above 2025's $201B. The models' middles all sit between $256B and $257B.

Reported profit carries the $9.9B state attorneys general settlement charge, which cuts the year's operating margin to about 31%, against 41% in 2025. Adjusted earnings add back $3.57 per share for that charge, giving about $32.24.

Adjusted earnings still include the $8.03B first-quarter tax benefit, because only the settlement charge is excluded. That lifts 2026 and makes 2027 look weaker by comparison.

Capital spending of about $137B, against $69.7B in 2025, leaves free cash flow near −$1.0B. The company narrowed its spending outlook to $135B to $145B, including finance leases.

With three quarters reported, ranges are narrow; the spread on earnings comes from fourth-quarter costs, tax and any new legal charge.

About $32.24 per share: nine months of reported earnings, plus $3.57 for the settlement charge, plus a fourth quarter near $8.23. The first-quarter tax benefit stays in.

#### Sales by quarter

Q2 2025 $47.5B +22%

Q3 2025 $51.2B +26%

Q4 2025 $59.9B +24%

Q1 2026 $56.3B +33%

Q2 2026 $60.8B +28%

Q3 2026 $64.6B +26%

Q4 2026 $75.0B +25%

Each percent is on the same quarter a year earlier. The forecast rows are the models' average.

#### What it rests on

**Settlement charge add-back**

$3.57 per share; third-quarter earnings $7.40 without it “•Diluted EPS of $3.83 would have increased by $3.57 to $7.40.” Third-quarter 2026 earnings release

**Capital spending outlook**

$135B to $145B including finance-lease payments, narrowed from $130B to $145B “We anticipate 2026 capital expenditures, including principal payments on finance leases, to be in the range of $135-145 billion, narrowed from our prior outlook of $130-145 billion.” Third-quarter 2026 earnings release

**First-quarter tax benefit**

$8.03B benefit, partly offsetting a 2025 tax charge, stays in adjusted earnings “Includes an $8.03 billion income tax benefit recognized in the first quarter of 2026, which partially offsets the $15.93 billion non-cash tax charge recorded in the third quarter of 2025 upon enactment of the One Big Beautiful Bill Act.” First-quarter 2026 earnings release

**Reality Labs losses**

Expected at $18B to $19B in 2026, below 2025 “At a segment level, we now expect Reality Labs losses to decline compared to 2025 to a range of $18-19 billion.” Third-quarter 2026 earnings release

**Full-year expenses**

$175B to $179B including the legal charge “We now expect full year 2026 total expenses to be in the range of $175-179 billion.” Third-quarter 2026 earnings release

#### Low and high

Base the middle forecast

Key assumptions

- ****Fourth-quarter sales near $75.0B take the year to about $257B.** Nine months brought $182B; the company guides $73B to $76B for the quarter. *3 of 6***
  - **GLM 5.3** gives about $256B for FY2026
  - **Grok 4.7** full-year 2026 sales are about 257 billion dollars
  - **Gemini 4 Argon** bringing full-year 2026 sales to about 256.7 billion dollars
- ****Only the $9.9B settlement charge is removed; the tax benefit and other charges stay.** The charge cut third-quarter earnings by $3.57 per share. *3 of 6***
  - **GPT-6.1 Sol** Only the attorneys general charge is removed.
  - **Grok 4.7** Adjusted earnings add that after-tax amount, about 9.2 billion dollars, back to 2026 net income and to no other year.
  - **GLM 5.3** I do not exclude the Q1 tax benefit or Q2's severance and legal charges from adjusted EPS
- ****Expenses near the middle of $175B to $179B; reported operating margin about 31%.** The guide includes the legal charge; nine-month costs were about $127B. *3 of 6***
  - **Grok 4.7** the full-year margin near 31 percent, because the legal charge stays in reported profit
  - **GPT-6.1 Sol** Fourth-quarter costs of $49.9 billion make full-year costs about $177 billion, inside the company's outlook.
  - **Opus 5.5** I used 49.3 billion, a little below the midpoint
- ****Capital spending about $137B of equipment purchases, leaving free cash flow near −$1.0B.** The $135B to $145B outlook includes about $4B of finance-lease payments not counted here. *3 of 6***
  - **Grok 4.7** That puts 2026 property purchases near 136 billion dollars and free cash flow near breakeven.
  - **GPT-6.1 Sol** I forecast $136 billion of property and equipment purchases in 2026
  - **Opus 5.5** so free cash flow is near zero

If all of this happens

**$257B** sales

**$32.24** adjusted EPS

**$137B** Capital spending

**−$1.04B** Free cash flow

Cash walk not supplied by every model.

Low about 1 chance in 10 of worse

What the low case assumes

- ****Fourth-quarter sales miss the bottom of the $73B to $76B guide.** The models' quarterly lows run down to $70.8B. *3 of 6***
  - **Grok 4.7** Capital spending stays at the top of the guided range while revenue misses the fourth-quarter guide
  - **GPT-6.1 Sol** a weak holiday outlook
  - **Muse Wonder** Fourth quarter revenue near or below the low end of company guidance
- ****A new legal charge is booked before year-end.** Extra settlement payments of up to $5.3B depend on YouTube and TikTok joining. *3 of 6***
  - **Grok 4.7** a new legal accrual of several billion dollars is booked
  - **Opus 5.5** further legal charges from youth trials
  - **Muse Wonder** new youth or other legal charges appear
- ****Fourth-quarter costs land at the top of what the expense guide allows.** The full-year guide leaves roughly $48B to $52B of fourth-quarter costs. *1 of 6***
  - **GPT-6.1 Sol** low earnings need cost and tax pressure as well as weak sales

If all of this happens

**$254B** sales · −1% vs base

**$30.75** EPS · −5% vs base

Earnings would have to fall to about $30.75, 4.6% below the base, on sales near $254B. That needs fourth-quarter sales below the guide or costs at its top, plus a new legal charge that adjusted earnings would not exclude. Grok goes furthest, to $29.50, because it combines a sales miss with capital spending at the top of the guide and a possible new accrual of several billion dollars.

Second-quarter costs already carried $2.4B of legal charges and $1.2B of severance, and adjusted earnings did not exclude them.

The six lows run from $253B (Grok 4.7) to $255B (Opus 5.5).

First sign A new legal accrual disclosed before December 31, 2026.

High about 1 chance in 10 of better

What the high case assumes

- ****Fourth-quarter sales beat the $76B top of the guide.** Third-quarter sales of $64.6B beat the top of the $61B to $64B range. *3 of 6***
  - **Grok 4.7** Fourth-quarter sales above 76 billion dollars
  - **Opus 5.5** Fourth quarter revenue above 76 billion dollars
  - **Gemini 4 Argon** fourth quarter 2026 revenue exceeding the 76 billion dollar top end of guidance
- ****Fourth-quarter costs come in below the expense guide's midpoint.** Headcount fell 15% from a year earlier after the May reduction. *2 of 6***
  - **Opus 5.5** I used 49.3 billion, a little below the midpoint
  - **Grok 4.7** spending stays at the low end of the guide
- ****Tax rate at the low end of 15% to 17%, and no new legal charge.** The company guides that range absent changes to its tax landscape. *1 of 6***
  - **Grok 4.7** The tax rate stays at the low end of the guided range and no further large legal accrual is booked

If all of this happens

**$259B** sales · +1% vs base

**$33.73** EPS · +5% vs base

Earnings would have to reach about $33.73, 4.6% above the base, on sales near $259B. Fourth-quarter sales would need to beat the top of the guide with costs and tax at the low end. Grok goes furthest, to $34.80, because it adds a tax rate at the low end of the guide and no further legal charge to a sales beat; Sol and Grok reach $260B on sales.

Third-quarter sales of $64.6B beat the top of the company's range, and operating income without the charge rose 11% from a year earlier.

The six highs run from $258B (GLM 5.3) to $260B (GPT-6.1 Sol).

First sign Holiday ad pricing through December 31, 2026, ahead of fourth-quarter sales above $76B.

#### Watch

Fourth-quarter sales above $76B and costs below the guide's midpoint point higher; a sales miss, costs at the top of the guide, or a new legal charge point lower.

##### Sales

PricedIn **$309B** $277B – $338B

$277B $309B $338B

##### adjusted EPS

PricedIn **$31.42** $22.83 – $39.67

$22.83 $31.42 $39.67

**Middle** the forecast **80% range** 1 in 10 chance below this bar, 1 in 10 above it. These are the low and high printed by each number.

Sales middles cluster from $305B to $316B. Muse, lowest, keeps new products small because 'paid consumer artificial intelligence products usually need several years to reach tens of billions of dollars of revenue.' Sol, highest on sales, adds 'roughly $4.5 billion and $14 billion from newer artificial intelligence products' yet is lowest on earnings, $29.02, with spending of $195B. Gemini, highest at $33.57, credits 'a fifteen percent leaner employee headcount'. Sol's earnings range, $19.00 to $43.00, is widest.

2027 sales of about $309B as growth slows, with earnings flat because depreciation and serving costs climb.

Sales **$309B** $277B – $338B · +20% on FY2026

Adjusted EPS **$31.42** $22.83 – $39.67 · −3% on FY2026

Base case

Operating margin **31.8%** 31.1% in FY2026

#### Investment and cash generation

Base case

Capital spending **$170B**

Free cash flow **−$14.0B**

Free cash flow: operating cash flow less capital spending.

The company gives no 2027 sales outlook but says its AI investments position it to sustain strong overall revenue growth throughout 2027. It expects Reality Labs losses to narrow to around $15B.

The models read that as growth of about 20%, to $309B, slowing from 28% in 2026 as comparisons get harder. Core advertising carries most of it; Muse, subscriptions and business agents add only a little in the middle.

Earnings per share average $31.42, 2.6% below 2026. Costs grow faster than sales as capacity built in 2026 and 2027 starts to depreciate, and 2026 earnings carried the one-time tax benefit.

Capital spending averages $170B, leaving free cash flow near −$14.0B. The company says its plans are geared to maximizing 2026 and 2027 capacity, and will give its 2027 expense and capital spending outlook at its next report.

The models split most on spending: Sol sees $195B and a margin near 29%; Grok sees $152B and a margin near 33%.

About $31.42 per share: sales near $309B, an operating margin near 32%, a tax rate near 16%, and a share count drifting up without buybacks, before the planned share dividend.

#### Sales by year

FY2023 $135B

FY2024 $165B +22%

FY2025 $201B +22%

FY2026 $257B +28%

FY2027 $309B +20%

FY2028 $361B +17%

The forecast years are the models' average.

#### Reality Labs losses: peak, then narrowing

2024 $17.7B

2025 $19.2B

2027 (company outlook) $15.0B

**Where each number comes from**

**2024**

“our investments in Reality Labs reduced our 2024 overall operating profit by approximately $17.73 billion”

**2025**

“our investments in Reality Labs reduced our 2025 overall operating profit by approximately $19.19 billion”

**2027 (company outlook)**

“Separately, we remain focused on improving Reality Labs profitability and expect losses to narrow next year to around $15 billion as we reallocate investment toward other priority areas, primarily AI.”

The hardware and virtual reality unit's losses peaked in 2025; the company expects them near $15B in 2027, which helps 2027 margins.

#### What it rests on

**2027 sales view**

No outlook given; positioned to sustain strong revenue growth throughout 2027 “Looking ahead, although we are not providing a full-year 2027 revenue outlook, the benefits our AI investments are delivering to the core business, combined with emerging opportunities from model advances and product launches, position us to sustain strong overall revenue growth throughout 2027.” Third-quarter 2026 earnings release

**Reality Labs losses**

Expected to narrow to around $15B in 2027 “Separately, we remain focused on improving Reality Labs profitability and expect losses to narrow next year to around $15 billion as we reallocate investment toward other priority areas, primarily AI.” Third-quarter 2026 earnings release

**Capacity plans**

Geared to maximizing 2026 and 2027 computing capacity “Consequently, our current plans are geared towards maximizing 2026 and 2027 capacity.” Second-quarter 2026 earnings call

**New data center leases**

About $68B signed in July 2026, starting in 2027 and 2028 “In July 2026, we entered into additional data center leases with lease obligations of approximately $68 billion, which are expected to commence in 2027 and 2028, with lease terms of 18 to 20 years.” Second-quarter 2026 quarterly report

**Server depreciation**

Most servers and network gear depreciated over 5.5 years “In January 2025, we completed an assessment of the useful lives of property and equipment, which resulted in an increase in the estimated useful lives of most servers and network assets to 5.5 years, effective January 1, 2025.” 2025 annual report

**Muse early adoption**

More than 10 million people set up Muse in six weeks “In just a month and a half, more than 10 million people in the U.S. and Canada have set up Muse personal agents.” Third-quarter 2026 earnings call

#### Low and high

Base the middle forecast

Key assumptions

- ****Sales grow about 20% to $309B as ad growth slows from the 2026 pace.** The company expects strong growth, but comparisons toughen and impression growth slowed to 10%. *5 of 6***
  - **Opus 5.5** set 2027 at about 19% (306 billion)
  - **Grok 4.7** The 2027 sales middle is 308 billion dollars, about 20 percent growth
  - **GLM 5.3** I take about 20% growth for 2027 (to roughly $308B)
  - **Gemini 4 Argon** roughly twenty-one percent in 2027 to 310 billion dollars
  - **Muse Wonder** I used about nineteen percent growth in the middle
- ****New AI products add only a few billion dollars in the middle.** Muse sign-ups and subscription trials are not yet paying customers. *4 of 6***
  - **Opus 5.5** I keep new AI revenue small in 2027 (a few billion dollars inside the 19% growth)
  - **GLM 5.3** I put only a modest contribution in my 2027–28 middles
  - **Muse Wonder** I therefore gave it only a small revenue contribution in 2027 in the middle
  - **GPT-6.1 Sol** a signup is not a paying customer
- ****Reality Labs losses narrow to around $15B.** The company said so, reallocating investment toward AI. *4 of 6***
  - **GLM 5.3** narrowing to about $15B in 2027
  - **GPT-6.1 Sol** I use around $15 billion of Reality Labs operating losses in 2027
  - **Muse Wonder** Reality Labs losses narrowing from 2026 to about fifteen billion dollars in 2027
  - **Grok 4.7** Reality Labs losses narrow to about 15 billion dollars next year
- ****Capital spending near $170B, so free cash flow stays near −$14.0B.** Plans maximize 2026 and 2027 capacity; leases signed in July start in 2027. *4 of 6***
  - **Opus 5.5** I set 2027 capital spending higher still, about 180 billion
  - **Muse Wonder** That supports very high capital spending in 2027
  - **GLM 5.3** capital spending keeps climbing into 2027 (~$160B in my base)
  - **Grok 4.7** Capital spending therefore steps up slightly in 2027, to about 152 billion dollars
- ****Costs grow about as fast as sales or faster; operating margin near 32%.** Capacity built in 2026 and 2027 depreciates over 5.5 years, and serving costs rise. *3 of 6***
  - **Opus 5.5** I set 2027 expenses up about 27% and margin about 31%
  - **GPT-6.1 Sol** operating margins are about 29% and 30%
  - **Grok 4.7** Operating profit in dollars grows only modestly after 2026 because depreciation and artificial intelligence serving costs rise almost as fast as sales
- ****No repeat of the 2026 tax benefit, which keeps earnings flat.** The $8.03B first-quarter 2026 benefit was one-time. *2 of 6***
  - **Grok 4.7** so 2027 earnings per share are about flat even though operating profit is a bit higher
  - **GPT-6.1 Sol** this explains why 2027 earnings can fall despite strong sales growth

If all of this happens

**$309B** sales

**$31.42** adjusted EPS

**$170B** Capital spending

**−$14.0B** Free cash flow

Cash walk not supplied by every model.

Low about 1 chance in 10 of worse

What the low case assumes

- ****Advertiser budgets shrink and sales growth falls to single digits.** Brent oil near $127 and weak consumer sentiment weigh on consumer brands' budgets. *3 of 6***
  - **Grok 4.7** Sales growth would fall toward the low single digits
  - **Opus 5.5** sales grow only about 7-8% a year
  - **GLM 5.3** Macro advertising pullback from 5-percent-plus Treasury yields and doubled energy prices
- ****Europe imposes further limits on personalized ads.** The company warns changes during its appeal could significantly hurt European revenue. *3 of 6***
  - **GLM 5.3** Further forced changes to European ad personalization during the appeal
  - **Gemini 4 Argon** Stricter European Union regulatory remedies on less personalized advertisements
  - **Grok 4.7** Europe tightens limits on personalized ads further
- ****Depreciation and serving costs outrun sales; margin falls toward the mid-20s.** Cloud contracts and data center leases already signed lock in rising costs. *3 of 6***
  - **Opus 5.5** operating margin falls toward the low-to-mid 20s
  - **Gemini 4 Argon** compressing operating margins to the mid-twenty percent range
  - **Muse Wonder** Infrastructure and legal costs rise faster than revenue
- ****A new legal charge or a share sale hits earnings per share.** Youth-related trials are scheduled, with claims up to the high tens of billions. *3 of 6***
  - **Opus 5.5** dilution if the new non-voting shares are sold to raise capital
  - **GPT-6.1 Sol** Additional legal expense or dilution can worsen earnings
  - **GLM 5.3** A new material legal charge from the remaining youth litigation

If all of this happens

**$277B** sales · −10% vs base

**$22.83** EPS · −27% vs base

Sales would have to fall to about $277B, 10% below the base, with earnings near $22.83. Ad budgets would shrink under high oil prices and weak consumer mood, Europe would tighten ad rules, and depreciation would land in full. Grok goes furthest on sales, to $264B, with growth falling toward the low single digits; Sol goes furthest on earnings, to $19.00, because its spending path is the heaviest and legal costs or new shares could add pressure.

Oil prices and interest rates are high, yet third-quarter sales still grew 26%; advertiser spending did dip when the Iran conflict began.

The six lows run from $264B (Grok 4.7) to $290B (GLM 5.3).

First sign A first-quarter 2027 outlook implying growth under 15% at the company's next report.

High about 1 chance in 10 of better

What the high case assumes

- ****Ad growth stays in the high 20s through 2027.** The fourth-quarter guide implies growth still close to the 2026 pace. *3 of 6***
  - **Opus 5.5** The ad engine keeps growing in the high-20s through 2027 and 2028
  - **Grok 4.7** Ad impressions and prices both keep growing at double-digit rates through 2027
  - **Gemini 4 Argon** drive advertising revenue growth above twenty-five percent in 2027
- ****Paid agents and subscriptions scale fast.** Subscriptions and trials passed 15 million; business-agent pricing began in August 2026. *3 of 6***
  - **GPT-6.1 Sol** Rapid paid adoption of consumer and business agents
  - **GLM 5.3** MetaOne subscriptions past 15 million and climbing quickly
  - **Muse Wonder** Muse, business agents and model services become material paid revenue in 2027 and 2028
- ****Compute efficiency and a smaller workforce hold cost growth below sales growth.** The company expects more than tenfold compute cost-efficiency gains over the next year. *3 of 6***
  - **GPT-6.1 Sol** Better computing efficiency and utilization
  - **Opus 5.5** compute efficiency gains
  - **Gemini 4 Argon** Tenfold improvements in artificial intelligence inference efficiency

If all of this happens

**$338B** sales · +9% vs base

**$39.67** EPS · +26% vs base

Sales would have to reach about $338B, 9.5% above the base, with earnings near $39.67. Ad growth would hold in the high 20s while Muse, subscriptions and business agents add billions, and compute efficiency keeps costs below sales growth. Sol goes furthest on sales, to $356B, because it lets paid agents scale fast; Sol and Grok reach $43.00 on earnings.

Automated Advantage Plus ad tools passed a $75B annual revenue run rate in the second quarter of 2026.

The six highs run from $328B (GLM 5.3) to $356B (GPT-6.1 Sol).

First sign A first-quarter 2027 outlook implying growth of 26% or more at the company's next report.

#### Watch

The 2027 expense and capital spending outlook at the next report: expense growth well above 27% lowers earnings, and spending far above $170B squeezes cash. A first-quarter outlook above 20% growth, or paid Muse revenue, points higher.

##### Sales

PricedIn **$361B** $300B – $421B

$300B $361B $421B

##### adjusted EPS

PricedIn **$37.05** $23.25 – $51.50

$23.25 $37.05 $51.50

**Middle** the forecast **80% range** 1 in 10 chance below this bar, 1 in 10 above it. These are the low and high printed by each number.

Sales middles run from $352B to $386B. Opus, lowest, slows growth to about 15% because 'growth usually slows as comparisons get harder after an acceleration year.' Sol, highest, assumes 'Advertising growth of about 21% in 2027 and 18% in 2028', but $225B of spending makes it lowest on earnings, $34.95. Gemini, highest at $40.85, has spending peak 'before moderating in 2028'. Sol's earnings range, $17.00 to $64.00, is far wider: its low needs weak sales plus cost pressure, its high strong sales plus cost discipline.

2028 sales of about $361B as growth slows to the mid-teens and earnings recover with steadier margins.

Sales **$361B** $300B – $421B · +17% on FY2027

Adjusted EPS **$37.05** $23.25 – $51.50 · +18% on FY2027

Base case

Operating margin **32.7%** 31.8% in FY2027

#### Investment and cash generation

Base case

Capital spending **$178B**

Free cash flow **$7.50B**

Free cash flow: operating cash flow less capital spending.

Neither the company nor the models have firm numbers for 2028. The company keeps 2028 capacity flexible, deciding server purchases closer to the time, and expects Reality Labs losses to keep narrowing.

The models expect growth of about 17%, to $361B, as the advertising base grows larger and comparisons toughen. Middles run from $352B for Opus to $386B for Sol.

Earnings per share recover to about $37.05, 18% above 2027, as operating margin edges up to about 33% and depreciation growth slows.

Spending is the widest split: Grok sees capital spending of $142B and free cash flow of $48.0B; Sol sees $225B and −$43.0B. The averages are $178B and $7.5B.

Ranges are wide two years out: on average, sales from $300B to $421B and earnings from $23.25 to $51.50.

About $37.05 per share: sales near $361B, an operating margin near 33%, a tax rate near 16%, and a share count drifting up about 1% a year, before the planned share dividend.

#### Each model's growth in FY2028

Opus 5.5 +15%

GPT-6.1 Sol +22%

Grok 4.7 +15%

GLM 5.3 +15%

Gemini 4 Argon +18%

Muse Wonder +16%

Growth of each model's middle on its own forecast for the year before.

#### What it rests on

**2028 capacity flexibility**

Plans maximize 2026 and 2027 capacity and keep flexibility for 2028 “I alluded in my main remarks to our focus on being on gearing our current infrastructure plans towards maximizing capacity in 2026 and 27, and giving us the flexibility to continue to grow in 28 beyond and beyond” Chief financial officer on a 2026 call

**Profit framework**

Grow operating profit over time rather than target a margin “We continue to use our framework that we shared, at this point several years ago of growing consolidated operating profit over time to guide those investments.” Fourth-quarter 2025 earnings call

**Muse profitability**

Company expects Muse highly profitable over several years if efficiency keeps improving “If compute efficiency continues improving quickly for another few years, which we think is highly likely, then Muse will be highly profitable over multiple years.” Third-quarter 2026 earnings call

**Legal exposure**

Damages sought across cases could total hundreds of billions of dollars “The maximum aggregate monetary damages or penalties sought across our various legal proceedings could amount to an aggregate of up to hundreds of billions of dollars and, as a result, could be material to the financial condition of the company.” 2025 annual report

#### Low and high

Base the middle forecast

Key assumptions

- ****Sales grow about 17% to $361B as growth slows toward the mid-teens.** A larger ad base and harder comparisons slow growth absent a large new revenue line. *5 of 6***
  - **Opus 5.5** In 2028 growth slows to about 15%
  - **Grok 4.7** 2028 steps down to the mid-teens
  - **GLM 5.3** For 2028 I assume mid-teens growth to about $355B
  - **Gemini 4 Argon** eighteen percent in 2028 to 365 billion dollars
  - **GPT-6.1 Sol** Advertising growth of about 21% in 2027 and 18% in 2028
- ****Operating margin edges up to about 33% as Reality Labs losses fall further.** Spending shifts from Reality Labs to AI, and depreciation growth eases after the 2026-2027 build. *4 of 6***
  - **Muse Wonder** recover in 2028 as Reality Labs losses fall and serving efficiency improves
  - **GLM 5.3** the margin holds around 32-33 percent
  - **Opus 5.5** In 2028 growth slows to about 15% and margin recovers slightly
  - **Grok 4.7** margins stay in the low-to-mid 30s in 2027 and 2028
- ****Capital spending around $178B; free cash flow near $7.5B.** The company keeps 2028 server purchases flexible rather than committed. *4 of 6***
  - **Grok 4.7** eases to about 142 billion dollars in 2028 rather than falling sharply
  - **Opus 5.5** For 2028 the company stresses flexibility, so I held spending roughly flat at about 185 billion
  - **Muse Wonder** a slower rise in 2028
  - **GLM 5.3** stays elevated in 2028 (~$175B)
- ****New AI products grow but stay small next to advertising.** Muse is costly to run until its compute efficiency improves. *3 of 6***
  - **GPT-6.1 Sol** Paid messaging, subscriptions and agents gradually broaden sales, without overtaking advertising.
  - **Muse Wonder** kept new product revenue modest in the middle
  - **GLM 5.3** a few points of growth
- ****Share count drifts up about 1% a year with buybacks paused.** No shares were repurchased in the first nine months of 2026. *3 of 6***
  - **Grok 4.7** equity grants typically add about 1 percent a year to the diluted share count
  - **GPT-6.1 Sol** Modest compensation dilution takes average diluted shares to about 2.62 billion in 2027 and 2.66 billion in 2028
  - **Opus 5.5** a flat-to-rising share count with no buybacks

If all of this happens

**$361B** sales

**$37.05** adjusted EPS

**$178B** Capital spending

**$7.50B** Free cash flow

Cash walk not supplied by every model.

Low about 1 chance in 10 of worse

What the low case assumes

- ****Sales growth slides toward single digits for two years.** Ad impression growth slowed from 19% to 10% across 2026. *3 of 6***
  - **Opus 5.5** sales grow only about 7-8% a year
  - **Grok 4.7** Sales growth would fall toward the low single digits
  - **GLM 5.3** Revenue growth slides to the low teens by 2028
- ****Further legal charges from youth trials hit earnings.** School district and individual cases remain after the attorneys general settlement. *3 of 6***
  - **GLM 5.3** A new material legal charge from the remaining youth litigation
  - **Gemini 4 Argon** potential additional legal accruals from youth bellwether trials
  - **Opus 5.5** further legal charges from youth trials
- ****Committed cloud and depreciation costs keep rising while sales slow.** Signed leases and cloud contracts run through 2028 and beyond. *3 of 6***
  - **Gemini 4 Argon** binding multi-year cloud commitments, high server depreciation
  - **GPT-6.1 Sol** Computing costs, legal exposure and financing pressure
  - **Muse Wonder** Infrastructure and legal costs rise faster than revenue

If all of this happens

**$300B** sales · −17% vs base

**$23.25** EPS · −37% vs base

Sales would have to fall to about $300B, 17% below the base, with earnings near $23.25. Ad growth would slide toward single digits for two years under a weak economy and European rules, while depreciation and cloud commitments keep costs rising. Grok goes furthest on sales, to $272B, with growth toward the low single digits; Sol goes furthest on earnings, to $17.00, because its base already carries $225B of spending and its low adds cost, tax and financing pressure.

The company's 2025 annual report warns that damages sought across its cases could total hundreds of billions of dollars.

The six lows run from $272B (Grok 4.7) to $325B (GLM 5.3).

First sign 2027 expense guidance at the company's next report implying costs growing far faster than sales.

High about 1 chance in 10 of better

What the high case assumes

- ****Muse, business agents and model services become material paid revenue.** At least 500 million small businesses use the apps; agents could transact directly. *3 of 6***
  - **Muse Wonder** Muse, business agents and model services become material paid revenue in 2027 and 2028
  - **GPT-6.1 Sol** Rapid paid adoption of consumer and business agents
  - **GLM 5.3** subscriptions, commerce fees and API revenue become material by 2028
- ****Compute efficiency bends the cost curve faster than expected.** The company expects more than tenfold compute cost-efficiency gains over the next year. *3 of 6***
  - **GPT-6.1 Sol** Better computing efficiency and utilization
  - **Gemini 4 Argon** Tenfold improvements in artificial intelligence inference efficiency
  - **Opus 5.5** compute efficiency gains
- ****Reality Labs losses narrow faster than the company's path.** About 70% of Reality Labs spending now goes to wearables, where glasses sales are growing. *3 of 6***
  - **GPT-6.1 Sol** Faster improvement in Reality Labs losses
  - **Muse Wonder** Reality Labs losses narrow quickly
  - **GLM 5.3** Reality Labs approaches break-even on glasses and subscriptions
- ****Ad growth holds above 20% through 2028.** Price per ad rose 14% in the third quarter, driven by ad performance gains. *2 of 6***
  - **Opus 5.5** The ad engine keeps growing in the high-20s through 2027 and 2028
  - **GLM 5.3** the ad business holds 20-percent-plus growth through 2028

If all of this happens

**$421B** sales · +16% vs base

**$51.50** EPS · +39% vs base

Sales would have to reach about $421B, 16% above the base, with earnings near $51.50. Ad growth would stay in the high 20s for a second year while Muse, business agents and model services add billions, and margin rises toward the mid-30s. Sol goes furthest, to $465B and $64.00, because it pairs the fastest paid-agent adoption with cost discipline.

The company expects Muse to become highly profitable over several years if compute efficiency keeps improving quickly.

The six highs run from $390B (GLM 5.3) to $465B (GPT-6.1 Sol).

First sign Paid Muse or business-agent revenue disclosed as a material line in a 2027 quarterly report.

#### Watch

The 2027 capital spending plan sets 2028 depreciation. Ad growth holding above 20% through 2027, or material paid agent revenue, points higher; growth sliding toward single digits or new legal charges points lower.

The years at a glance · sales

FY2023 **$135B**

FY2024 **$165B**

FY2025 **$201B**

FY2026 **$257B** $254B – $259B

FY2027 **$309B** $277B – $338B

FY2028 **$361B** $300B – $421B

$200B $300B $400B

The years at a glance · adjusted EPS

FY2023 **$14.87**

FY2024 **$23.86**

FY2025 **$23.49**

FY2026 **$32.24** $30.75 – $33.73

FY2027 **$31.42** $22.83 – $39.67

FY2028 **$37.05** $23.25 – $51.50

$20.00 $30.00 $40.00 $50.00