Tesla’s “Project Crystal Sun” is a real tax-incentive filing for a $10.1 billion vertically integrated solar-cell and module plant in Fort Bend County, Texas. It is not yet a committed factory.

Commercial production is targeted for the first quarter of 2029 if Tesla picks the Texas site, wins incentives, and clears permits. Tesla has not disclosed annual gigawatt output, so any panel-count figure is an estimate, not an official number.

The internal name is Project Crystal Sun. The public paper trail is a Jobs, Energy, Technology and Innovation (JETI) application, file J0050, with Lamar Consolidated ISD and the Texas Comptroller of Public Accounts. It was signed July 22, 2026, posted August 6, and later supplemented on August 14 and August 20.

Tesla has not issued a standalone press release announcing the plant. A company spokesperson did not comment when asked. The closest official Tesla language is in the Q2 2026 shareholder update and Form 10-Q: site work and equipment procurement for solar manufacturing “progressed,” and the company is “developing our solar manufacturing.” CFO Vaibhav Taneja said on the earnings call that Tesla intends to multiply U.S. solar manufacturing “by an order of magnitude.”

What Tesla actually put on paper

The application describes a solar-cell manufacturing campus on portions of five parcels totaling about 3,050 acres near Richmond / Needville, along FM 762 and FM 1994 in Fort Bend County, southwest of Houston.

Planned investment: $10.116 billionAbout $1.5 billion real property
About $8.6 billion equipment and personal property

Spend is front-loaded in three years:2026: about $2.33 billion
2027: about $3.29 billion
2028: about $4.49 billion

Then commercial operations in Q1 2029. The requested JETI limitation would run 2029–2038.

The equipment list is the important part. Tesla is not describing a simple module-assembly line like Buffalo. The filing lists ingot manufacturing, wafer manufacturing, coating, metallization and printing, cell testing, cleanrooms, and automated material handling — an ingot-to-cell (and potentially module) stack. The NAICS code is 334413, semiconductor and related device manufacturing.

Kroll’s economic statement attached to the filing projects 9,712 permanent jobs at full buildout, a peak of about 1,147 construction jobs, roughly $107 billion in Texas GDP impact, and about $6.4 billion in state and local tax revenue over 38 years. Those are consultant projections used to support a tax deal, not audited results.

How real is it?

The filing is real. The factory is not locked.

Tesla’s own application says it is evaluating sites “across multiple U.S. states,” comparing Fort Bend with an unnamed out-of-state alternative, and that without the JETI limitation plus local abatements the Texas site’s economics are worse. If incentives fail, Tesla says it would further evaluate the investment outside Texas.

• Other limits:No federal, state, or local construction permits have been filed for this site, according to reporting on the paperwork.
• The required reinvestment zone is listed as “TBD.”
• No ground has been broken.
• The 9,712-job headline is a 2033 full-staff figure, not opening-day employment. Reporting on the job schedule shows commercial operations starting in 2029 with about 1,942 workers, then adding a similar increment each year through 2033.

That is how large Texas incentive filings work: they put a maximum project on paper so the company can shop tax treatment. SpaceX used the same JETI process for its Texas chip plans. Crystal Sun should be treated as a site-selection and incentive bid, backed by real spending signals, not as a notice to proceed.

Those spending signals matter. Tesla’s Q2 2026 materials already listed solar manufacturing among multi-year infrastructure buildouts. Earlier in 2026, reporting pointed to large purchases of solar manufacturing equipment (including from Chinese toolmakers such as Suzhou Maxwell) and to a possible Houston-area solar build near Tesla’s Megapack Megafactory in Brookshire. Crystal Sun is a different Fort Bend greenfield, about 45 minutes southwest of Houston. It may replace, complement, or compete with that earlier site rumor. Tesla has not publicly reconciled the two.

Elon Musk’s January 2026 World Economic Forum remark is the strategic frame Tesla cited in the filing: SpaceX and Tesla teams are separately working toward 100 GW a year of manufactured solar power in the United States, “that’ll probably take us three years or something.” A Tesla solar-manufacturing job posting earlier in 2026 used similar language: a 100 GW facility from raw materials in the United States before the end of 2028. Crystal Sun is the first large public capital number attached to that goal. It is not proof the 100 GW target will be hit on that timetable.

How fast it can come online, and when it starts

On Tesla’s own calendar:

• Construction/capital spend: 2026 through 2028
• First commercial production: Q1 2029
• Full employment: 2033 under the filing’s ramp

“This year” construction is possible only if Texas wins the site fight and local approvals move fast. As of late August 2026, the Comptroller page still shows the application in the posted / supplement phase, not an executed JETI agreement. After completeness review, the Comptroller has a statutory window to recommend the project to the Governor; the school district and Governor must then agree. Fort Bend County would still need to create the reinvestment zone.

A 2029 start is aggressive for a greenfield, vertically integrated cell fab, but it is consistent with buying turnkey Asian production tools and running Tesla as its own general contractor — something Taneja noted Tesla already does on most of its factories. It is not consistent with Musk’s “about three years” 100 GW comment if that clock started at Davos in January 2026. A plant that first ships in early 2029 cannot be at 100 GW in 2028. Investors should treat 100 GW by 2028 as an aspiration across Tesla’s whole solar program (Buffalo plus new sites), not as Crystal Sun’s opening rate.

How many panels can it build in a year?

Tesla did not disclose nameplate capacity. That is the most important capacity fact in the filing.

What is known:

Buffalo / Gigafactory New York currently assembles Tesla’s new TSP-420 residential panels at on the order of 300 MW per year — Tesla’s first in-house-designed conventional module, shipped starting Q1 2026. Crystal Sun would be a different order of magnitude if built as described.

Musk’s public target is 100 GW/year of U.S. manufactured solar for Tesla (and separately for SpaceX).
One industry read of the $10.1 billion capex is that this campus is a foundation stone toward that target, not the entire 100 GW by itself. Another estimate circulating in trade coverage suggested on the order of $16 billion might be needed to reach 100 GW of fully integrated U.S. capacity. Neither figure is Tesla guidance.

A simple illustration, not a forecast: 100 GW is 100 billion watts. At Tesla’s current 420 W residential module, that would be roughly 238 million panels a year. Utility modules in the 600–700 W class would mean fewer physical panels for the same watts. Crystal Sun’s actual annual output could be a slice of that, all of it, or something else entirely. Until Tesla names a GW rating, panel-count headlines are speculation.

Homeowners and edge-grid resilience, or solar farms and Megapacks?Both, according to the filing. Storage pairing is Tesla’s business model, not a stated exclusive offtake contract for this plant.

The application says the plant would make photovoltaic cells and/or assembled modules “for deployment in utility-scale, commercial, and distributed solar installations,” and discusses utility-scale plus on-site, behind-the-meter generation. That is the full market stack: rooftops, C&I, and solar farms.

Tesla’s current consumer product is already a resilience package. Official Tesla materials describe 420 W panels engineered in California, assembled in Buffalo, with 18 “power zones” for shade performance, a rail-less Panel Mount, and integration with Powerwall and the Tesla app. Tesla discontinued Solar Roof tiles in August 2026 and pointed customers to conventional panels. Energy revenue growth has been storage-led — Powerwall and Megapack — not rooftop generation. Megafactory Texas is scheduled to start Megapack production in 2026, which would sit in the same broader Houston-area energy cluster if Crystal Sun is built.

The strategic logic is obvious: domestic cells reduce tariff and supply-chain risk; Tesla can feed Buffalo residential assembly, third-party installers, its own storage-plus-solar packages, and utility projects that sit in front of Megapacks. Musk has also tied large-scale solar to AI electricity demand and, separately through SpaceX, to space-based power. The Fort Bend filing does not assign output to orbital arrays. SpaceX has its own solar-manufacturing track.

For Energy News Beat readers, the clean read is this: Crystal Sun is designed as a cell and module foundry, not a residential-only factory and not a captive solar-farm EPC shop. If Tesla executes, homeowners get a more secure domestic panel supply paired with Powerwall. Utilities and data-center offtakers get a potential U.S. cell source that can be coupled with Megapack. The mix will be set by price, 45X production credits, and which customer can absorb volume first.

What investors should watch

• Site decision, not the headline capex. Watch for a Comptroller recommendation, Governor and Lamar CISD action, and a Fort Bend reinvestment zone. A competing-state win would make Crystal Sun a footnote.
• Permits and dirt. The first hard confirmation is a site plan, air/water permits, and construction photos — not another incentive PDF.
• Named capacity. The first GW figure in an earnings deck, 10-K, or job posting that is tied to Texas is more valuable than the $10.1 billion number.
• Federal stack. The filing flags IRA-era Section 45X (advanced manufacturing production) and Section 48D (advanced manufacturing investment) support. Policy risk on those credits is project risk.
• Job and spend ramp vs. 2029 ops. Opening-year jobs near 2,000 and heavy 2026–28 equipment receipts would corroborate the timeline. A 9,712-job boast in 2026 does not.
• Energy-segment mix. Storage is carrying Tesla Energy today. Watch whether generation deployments and solar ASPs reappear in disclosures as Buffalo ramps and if Texas cells are mentioned as a cost or tariff hedge.
• Integration with Megapack Texas and Buffalo. Cells shipped to New York versus full modules made in Fort Bend tells you whether Tesla is building a U.S. cell champion or a closed rooftop-to-farm product line.
• Execution versus rhetoric. Musk’s 100 GW / ~2028 language and a Q1 2029 commercial start are not the same plan. Slippage on tools, polysilicon, water, or labor would show up first in capex timing, not in a press release.

Crystal Sun is the largest U.S. manufacturing investment Tesla has put on a Texas incentive form. It is also still a maybe. The factory becomes “real” when Tesla stops evaluating other states and starts pouring slabs.